The use of capnography for real-time monitoring of mask ventilation during induction of general anaesthesia 2023

Editor,

Effective mask ventilation with proper airway manoeuvres secures maintenance of pulmonary ventilation and oxygenation during induction of general anaesthesia.¹ However, life-threatening difficulties and impossible mask ventilation do occur. The difficult airway guidelines of the American Society of Anesthesiologists and the Difficult Airway Society do not propose any objective criteria for diagnosis of difficult and impossible mask ventilation.^2,3 Therefore, real-time breath-by-breath monitoring of mask ventilation effectiveness could maximise the risk management strategies of difficult airway guidelines.^2–5

Capnography monitoring is routinely used during anaesthesia induction for confirmation of tracheal intubation. The rising trace of the capnogram (phase II) reflects the change in the expired gas composition because of mixing of the dead space (anatomical and physiological) gas and alveolar CO₂. Furthermore, the end-tidal CO₂ concentration (ET CO₂) during phase II theoretically changes from zero to alveolar CO₂ concentration, which may possibly reflect the tidal volume.⁶

The aim of this prospective observational study was to test the hypothesis that tidal volume during mask ventilation can be estimated by the combination of the capnogram waveform and ET CO_2, enabling the determination of threshold ET CO₂ values, which indicate critical and secure ventilation, defined as a tidal volume less than 2 ml kg⁻¹ and greater than 5 ml kg⁻¹ ideal body weight (IBW).

Following ethical approval (Ethical Committee of Graduate School of Medicine, Chiba University, Chiba, Japan; number 3203) and clinical study registration (UMIN000035108, 3 December 2018: https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000039815), written informed consent was obtained from each patient. Between 3 December 2018 and 21 January 2020, adult patients scheduled for surgery under general anaesthesia were screened. The following patients were excluded: candidates for awake intubation, full stomach, undergoing emergency surgery, tracheostomy and tracheoesophageal fistula, dementia and with allergy for rocuronium or propofol.

After loss of consciousness, rocuronium 0.7 mg kg⁻¹ was injected. The anaesthetist performed one handed mask ventilation through an anaesthesia full-face mask for 1 min as usual but was instructed not to exceed a peak airway pressure of 20 cmH₂O. Twelve anaesthesia providers with a wide range of experience participated in this study without any prior knowledge of the purpose of the study.

Expiratory tidal volume (GF-220R Multigas/Flow unit, Nihon Kohden, Tokyo, Japan), and ET CO₂ concentration with a mainstream capnometer (CAP-ONE, TG-970P, Nihon Kohden, Tokyo, Japan, accuracy: ±2 mmHg) were continuously measured. Three distinct capnography waveforms were identified: plateaued capnogram, upswing capnogram (lack of the phase III plateau) and flat-line capnogram (ET CO₂ zero). The monitoring screen and video images of the patient’s head, face and neck, and the operating room were integrated synchronously on the same screen, and these were recorded to a hard disk for later detailed breath-by-breath analyses of the ventilation variables.

On the basis of our previous study,⁷ the sample size was determined to be 52 breaths for each upswing capnogram category assuming α = 0.05 (two tailed), β = 0.8 to detect a clinically meaningful ET CO₂ difference of 4 mmHg. We theorised that 50 patients could produce more than 52 breathes for each capnogram category during anaesthesia induction.

Two-thirds of all upswing capnogram breaths were randomly selected, and receiver operating characteristic (ROC) curves and optimal cutoff values were estimated. The remaining one-third of traces were used as a validation dataset for the cutoff value, and we calculated sensitivity, specificity and AUC. A value of P less than 0.05 was considered as statistically significant, and all P values were two-sided. All statistical analyses were performed using SAS Ver. 9.4 (SAS Institute, Cary, North Carolina, USA).

From a total of 1219 breaths in 50 participants (predominantly nonobese females), capnography waveform classification was possible in 1215 breaths: these included 920 (76%) plateaued, 244 (20%) upswing and 51 (4%) flat-line capnograms. A plateaued capnogram had moderate diagnostic performance of secure ventilation with 85% sensitivity and 81% specificity. A flat-line capnogram had a highly specific diagnostic performance of critical hypoventilation with 60% sensitivity and 99% specificity. In contrast, an upswing capnogram alone had a poor diagnostic performance for both secure and critical ventilation.

From the 244 upswing capnogram breaths, 164 breaths and 80 breaths were randomly selected for ROC analyses and cross validation tests. Figure 1 presents the ROC analysis result for prediction of critical and secure ventilation in upswing capnogram breaths and determines the cutoff ET CO₂ values to be 10 and 22 mmHg with AUC values of 0.927 and 0.907, respectively. Using the cutoff values, cross validation test resulted in 82% sensitivity and 67% specificity for critical hypoventilation prediction, and 67% sensitivity and 93% specificity for secure ventilation prediction.

Although tidal volume can be measured directly by various monitors, a threshold for critical hypoventilation volume should be individualised by the patient’s body size: the difficulty in such tidal volume normalisation may limit the clinical usefulness of monitoring absolute tidal volume for on-site decision making. This study is the first to systematically test the performance of the capnogram and ET CO₂ for monitoring the effectiveness of mask ventilation. Unlike tidal volume, the normal ET CO₂ range, while depending on cardiac output and body temperature, is independent of age and body size.

On the basis of the findings of this study, we propose a simple decision-making algorithm by using the combination of the capnogram and ET CO₂ during induction of general anaesthesia. First, either a plateaued capnogram or an upswing capnogram with ET CO₂ greater than 22 mmHg guarantees secure mask ventilation with oxygenation. When the ET CO₂ in an upswing capnogram is between 10 and 22 mmHg, the anaesthesiologist should try to maintain ET CO₂ at above 22 mmHg by using advanced airway techniques such as insertion of an oral airway, and/or two-handed mask ventilation. If ET CO₂ cannot be increased by these interventions, the anaesthesiologist needs to consider the use of the difficult airway algorithm, including supraglottic airway insertion, wakening the patient and surgical cricothyroidotomy. When there is no capnogram waveform or ET CO₂ less than 10 mmHg in an upswing capnogram is persistent, then the difficult airway algorithm should be commenced immediately.

In conclusion, the combination of a capnogram and ET CO₂ concentration accurately estimated a normalised tidal volume. ET CO₂ greater than 22 mmHg is the minimum target for achieving well tolerated and effective mask ventilation during induction of general anaesthesia. The clinical usefulness of the proposed real-time mask ventilation grading system needs to be tested in future large-scale multicentre studies, including patients with difficult airway.