記住我
ORIGINAL ARTICLE
Year : 2021 | Volume
: 70
| Issue : 1 | Page : 81-88
Role of capnography in detecting hypercapnic events during weaning from mechanical ventilation
Mona Mansour1, Fatma A.M Elmekawy2, Haytham S Diab1
1 Department of Respiratory Medicine, Ain Shams University, Egypt
2 Respiratory Intensive Care Unit, Abbasia Chest Hospital, Cairo, Egypt
Correspondence Address:
MD Haytham S Diab
Department of Respiratory Medicine, Ain Shams University, Cairo, 11835
Egypt
Source of Support: None, Conflict of Interest: None
DOI: 10.4103/ejcdt.ejcdt_107_20
Background End-tidal carbon dioxide (ETCO2) measurement is more applied in clinical practice, as it is a noninvasive tool that helps to evaluate the ventilatory status of the patient. The aim of the present study was to investigate the relationship between ETCO2 and arterial partial pressure of CO2 (PaCO2) measurements and to estimate the role of capnography during weaning off mechanical ventilation.
Patients and methods A prospective cross-sectional study was conducted on 60 adult critically ill patients with respiratory failure, comprising 47 males and 13 females. They were put on mechanical ventilation, and after fulfilling criteria of weaning, they were kept on continuous positive airway pressure (CPAP) mode for 1 h. ETCO2, PaCO2, and gradient were measured before and after weaning.
Results Mean difference between PaCO2 and ETCO2 before weaning was 8.3 and after one hour on CPAP was 5.1. There was a highly statistically significant positive correlation between ETCO2 and PaCO2 before weaning and after 1 h on CPAP. Bland-Altman comparison between ETCO2 and PaCO2 before weaning and after 1 h on CPAP revealed limits of agreement were 1.941 to 14.826 and −4.134 to 14.367, respectively.
Conclusion ETCO2 measurement gives a proper estimation of PaCO2 in patients on mechanical ventilation. ETCO2 decreases the need for repeated invasive arterial blood gas samples.
Keywords: arterial partial pressure of CO2, capnography, end-tidal CO2, weaning from mechanical ventilation
Observation of alveolar ventilation is an important part of providing current ventilatory support to patients with acute and chronic respiratory failure.
Multiple arterial blood gases (ABG) are related with increased blood loss and catheter-induced sequalae in weaning patients in ICU, especially if ABG is taken by standard methods. Furthermore, ABG only gives a snapshot of ventilatory status and lacks data about the dynamic assessment of alveolar ventilation [1]. So, a continued, noninvasive surveillance strategy is surely preferred during the weaning trials [2].
Capnometry is the measurement of carbon dioxide in respiratory gases, whereas capnography is the continuous measurement of the partial pressure of CO2 (PaCO2) in respiratory gases [3].
Capnography provides a continuous, noninvasive measurement of CO2 in respiratory gases to evaluate ventilatory status [4], to ensure the position of the endotracheal tube, to follow the efficiency of cardiopulmonary resuscitation, and to identify the possible reasons of bronchospasm [5].
End-tidal carbon dioxide (ETCO2) can be measured by two techniques depending on the position of the sensor: sidestream and mainstream. In the mainstream technique, the sensor is placed directly in the patient’s breathing circuit. Mainstream units have been essentially evolved for intubated patients. The main privilege of this method is the immediate gas analysis. Mainstream units can be also used to estimate ETCO2 in compliant patients who can blow into the sample tube several times [6]. The sidestream technique can be used in both intubated and nonintubated patients, as expiratory air is passed through the sample tube to the sensor that is away from the patient’s airway. Moreover, the precision of sidestream method is decreased owing to increase in dead space resulting from suction catheters or obstructing the catheter by fluids and secretions that can sometimes cause inaccurate results [7] and measurement may take prolonged time. Interestingly, it is mentioned in several studies that in patients with ventilation/perfusion mismatch like emphysema, capnometer is better to be avoided occasionally owing to the concern that ETCO2 estimation may not reflect PaCO2 accurately [6].
There are certain conditions associated with lowering level of ETCO2 like cardiac arrest, partial airway occlusion, pulmonary embolism, hyperventilation, hypotension, hypothermia, hypovolemia, and leaks in sample tube. However, increased level of ETCO2 may happen in hypoventilation, fever, ventilation of a collapsed lobe or lung, and bronchospasm. Moreover, previous models of mainstream sensors use heat to decrease the condensation of water vapor to obviate wrong elevated levels of carbon dioxide, but the heat carries the hazard of burning the patient’s skin. Moreover, precaution must be considered to decrease the extra weight put on the airway to decrease the probability of unintended endotracheal extubation [8].
The aim of the present study was to investigate the relationship between ETCO2 and PaCO2 measurements and to estimate role of capnography during weaning from mechanical ventilation.
Patients and methodsStudy design
A prospective cross-sectional study was performed at the respiratory ICU of Abbasia Chest Hospital during the period from April 2018 to April 2019.
Study population
A total of 60 adult critically ill patients with respiratory failure, comprising 47 males and 13 females, were enrolled in the present study. They were admitted to the respiratory ICU and fulfilled the criteria of mechanical ventilation. The mean age was 60.5 (SD: 16.4, range: 19–95) years. Of them, 26 had chronic obstructive pulmonary disease (COPD), 10 pneumonias, seven tuberculosis, four congestive heart failure and pulmonary hypertension, four bronchiectasis, three interstitial lung disease, three obstructive sleep apnea, two bronchial asthmas, and one lung cancer. Tracheostomized patients and those who were extubated within the first 48 h of mechanical ventilation were excluded.
Methods
All the patients were orally intubated with a cuffed endotracheal tube with an internal diameter ranging from 7.5 to 8.5 mm. Patients were ventilated in assisted volume control mode or synchronized intermittent mandatory ventilation (SIMV) (MAQUET Servo, Maquet Critical Care AB, Solna, Sweden).
All the patients involved in the present study were assessed on admission by full history taking from patients’ relatives, clinical examination, routine laboratory investigations, plain chest radiography, ECG, and ABG samples, which were drawn by radial arterial puncture in heparin-rinsed plastic syringes and immediately analyzed for oxygen and carbon dioxide partial pressures (GEM Premier 3000; Instrumentation Laboratory Company, Bedford, Massachusetts, USA).
The patients were followed by monitors (Drager Medical System Inc., Telford, Pennsylvania, USA) with built-in pulse oximeter and CO2 analyzer. Mainstream capnography was connected on the expiratory side of the circuit’s endotracheal tube connector after proper calibration and adjustment. It was used to monitor continuously CO2 partial pressure changes against time to measure ETCO2, and gradient between PaCO2 and ETCO2 was determined.
After stabilizing the medical conditions of the patients and meeting the parameters for weaning off mechanical ventilation, ETCO2 was recorded simultaneously, whereas the arterial blood was taken in the syringe to evaluate the relationship between PaCO2 and ETCO2 before start weaning. All patients were placed on continuous positive airway pressure (CPAP) for a period of 1 h followed by reassessment of PaCO2 and ETCO2. Later on, the final outcome of the studied patients was identified, which resulted in 42 patients with successful weaning off mechanical ventilation and extubation and 18 patients with failed weaning and returned on assisted volume control mode. The study was approved by the Institutional Ethical Committee at March 2018.
Study end points
The primary end point was to evaluate the relationship between the change in PaCO2 and mainstream ETCO2 before and after a weaning trial. The second end point was to determine the ability of ETCO2 to identify clinically relevant episodes of hypercapnia during weaning from mechanical ventilation.
Statistical methods
Data were revised for completeness and consistency. Data entry was done on Microsoft Excel workbook (Microsoft, New York, NY 10022-4210, 677 Fifth Avenue, USA). Quantitative data were summarized by mean and SD, whereas qualitative data were summarized by frequencies and percentages. The program used for data analysis are IBM SPSS statistics for windows version 23 (IBM Corp., Armonk, New York, USA). χ2-test, Pearson correlation coefficient test, and Student t-test were used in the analysis of this paper. A ‘P value’ of less than 0.05 was considered statistically significant. Moreover, Bland-Altman graphs were used to test agreement between PaCO2 and ETCO2. STATA 10 Program (StataCorp. 2007, Stata Statistical Software: release 10; StataCorp LP, College Station, Texas, USA) was used to perform the Bland-Altman graphs and confidence limits.
ResultsA total of 60 patients were enrolled in this study, with mean age of 60.5 years (SD: 16.3; median: 64; range: 19–95). Of them, 47 (78.3%) were males. All the involved patients were mechanically ventilated owing to respiratory failure and were classified according to the final outcome from mechanical ventilation, as either successful weaning from mechanical ventilation [42 (70%) patients with mean age: 60.5 years] or failed weaning from mechanical ventilation [18 (30%) patients with mean age: 60.6 years; 95% confidence interval (CI): 18.8–43.2%], with no statistically significant difference between the two groups of patients regarding the mean age (P=0.9). Main characteristics of studied patients are shown in [Table 1]. Descriptive statistics of clinical parameters and laboratory investigations among the studied patients are shown in [Table 2].
Table 2 Descriptive statistics of clinical parameters and laboratory investigations among the studied patientsThere was a higher percentage of failure of weaning among nonmarried patients (47.3%) compared with among married patients (21.9%), and the difference was statistically significant, with odds ratio of 3.9; this means that failure of weaning was almost four times in the studied nonmarried patients compared with the studied married patients. However, there were no statistically significant relations between failure of weaning and following items: sex, residence, smoking, absence or presence of comorbidities with its number, and diagnosis, as shown in [Table 3].
Table 3 Comparison between sex, marital status, residence, diagnosis, and comorbidities among the studied patients and outcome of weaning from mechanical ventilationDescriptive statistics of blood gases on admission, before weaning, and after one hour on CPAP were noted. Moreover, descriptive statistics of ETCO2 before weaning and after 1 h on CPAP were noted, as shown in [Table 4], where the mean ETCO2 was lower than the mean PaCO2 before weaning and after 1 h on CPAP.
Table 4 Descriptive statistics of blood gases on admission, before weaning and after 1 h on CPAP and descriptive statistics of ETCO2 before weaning and after 1 h on CPAPThe mean difference between PaCO2 and ETCO2 among studied patients before weaning was 8.3±3.2 (minimum: 1 and maximum: 15) and after 1 h on CPAP was 5.1±4.6 (minimum: −15 and maximum: 16), as shown in [Table 5].
Table 5 The mean difference of PaCO2 compared to ETCO2 among studied patientsThere was a highly statistically significant positive correlation between PaC02 and ETCO2 before weaning and after 1 h on CPAP, as shown in [Table 6] and [Figure 1] and [Figure 2].
Table 6 Correlation coefficient between PaCO2 and ETCO2 before weaning and after 1 h on CPAP
Figure 1 Correlation between PaCO2 and ETCO2 among the studied patients before weaning. ETCO2; end-tidal carbon dioxide; PaCO2, partial pressure of CO2.
Figure 2 Correlation between PaCO2 and ETCO2 among the studied patients after 1 h on CPAP. CPAP, continuous positive airway pressure; ETCO2; end-tidal carbon dioxide; PaCO2, partial pressure of CO2.Bland-Altman comparison between PaCO2 and ETCO2 among the studied patients before weaning revealed that the limits of agreement were 1.941–14.826, the mean difference was 8.383 mmHg (CI: 7.551–9.215 mmHg), and the range was 23–77.5, as shown in [Figure 3].
Figure 3 Bland-Altman comparison PaCO2 and ETCO2 among the studied patients before weaning. Limits of agreement (reference range for difference): 1.941–14.826; mean difference: 8.383 (CI: 7.551–9.215; range: 23–77.5). CI, confidence interval; CPAP, continuous positive airway pressure; ETCO2; end-tidal carbon dioxide; PaCO2, partial pressure of CO2.Moreover, Bland-Altman comparison between PaCO2 and ETCO2 among the studied patients after 1 h on CPAP revealed that the limits of agreement were −4.134 to 14.367, the mean difference was 5.117 mmHg (CI: 3.922–6.311 mmHg), and the range was 25 500–84 000, as shown in [Figure 4].
Figure 4 Bland-Altman comparison of PaCO2 and ETCO2 among the studied patients after 1 h on CPAP. Limits of agreement (reference range for difference): −4.134 to 14.367; mean difference: 5.117 (CI: 3.922–6.311); and range: 25.500–84.000. CI, confidence interval; CPAP, continuous positive airway pressure; ETCO2; end-tidal carbon dioxide; PaCO2, partial pressure of CO2. Discussion ETCO2 estimations are influenced in general by PaCO2 levels, volume of dead space, and pulmonary vasculature. ETCO2 mainly relies on alveolar CO2 (PACO2). Nonhomogenous alveoli CO2 emptying patterns as occurred in patients with ventilation/perfusion mismatch will lead to mismatched PACO2 and underestimation of PaCO2 levels. A high ventilation/perfusion ratio and increased dead space volume lead to low ETCO2 levels in comparison with PaCO2, whereas a low ventilation/perfusion ratio has diminished effect on producing smaller ETCO2 levels in comparison with PaCO2[9].
Usually, there is a difference between PaCO2 and ETCO2 of approximately 2–5 mmHg in normal intubated patients, and this difference increases with age, pulmonary diseases, impairment of cardiac function, and others conditions that are commonly found in patients admitted to emergency department for respiratory distress [10].
The results of this study revealed that the mean ETCO2 was lower than the mean PaCO2 before weaning and after 1 h on CPAP. Moreover, there was a highly statistically significant positive correlation between PaC02 and mainstream ETCO2 before weaning and after 1 h on CPAP.
Furthermore, our results showed that the means of both PaCO2 and ETCO2 before weaning were less than means of them after 1 h on CPAP. These increments in both means of PaCO2 and ETCO2 from mechanical ventilation to CPAP could be referred to the relative hyperventilation achieved in mechanical ventilation that normalized on CPAP. However, mean difference between PaCO2 and ETCO2 before weaning was more than the mean difference of them after 1 h on CPAP.
Interestingly, mean difference between PaCO2 and ETCO2 before weaning was 8.3±3.2 mmHg and limits of agreement were 1.9–14.8, whereas after 1 h on CPAP, mean difference between PaCO2 and ETCO2 was 5.1±4.6 mmHg and limits of agreement were −4.134 to 14.367 by Bland-Altman comparison.
In accordance with our results, Saura et al. [11] enrolled 30 critically ill patients with different pulmonary diseases presented by respiratory failure and were put on mechanical ventilation and they were tried on CPAP for 2 h. They found a significant correlation (r=0.74; P<0.01) regarding changes in PaCO2 and ETCO2 between mechanical ventilation and the first and second hour on CPAP (we performed follow-up for our studied patients for only 1 h on CPAP). Moreover, our results agreed with Saura et al. [11], who revealed that means of both PaCO2 and ETCO2 before weaning were less than means of them after 1 h on CPAP in all studied patients, but unlike our results, mean difference between PaCO2 and ETCO2 before weaning and after 1 h on CPAP remained unchanged in all their studied patients.
Our results coincided with results of Razi et al. [12], who performed a cross-sectional study on 87 adult mechanically ventilated patients owing to different pulmonary diseases. They studied the relation between PaCO2 and mainstream ETCO2 in SIMV and CPAP modes. They found a highly statistically significant positive correlation between means of both mainstream ETCO2 and PaC02 before weaning on SIMV and on CPAP (r=0.893, P<0.0001 and r=0.841, P<0.0001, respectively). Moreover, they reported that the mean of ETCO2 was lower than the mean of PaCO2 in all modes. In addition, Razi et al. [12] stated that mean difference between PaCO2 and ETCO2 in SIMV (3.37±7.93, 95% CI: 1.77–4.97) was more than in CPAP (2.32±5.62, 95% CI: 0.98–3.67), which were matched with our results.
In contrast to our results, Russell and Graybeal [13] revealed in a study conducted on 11 critically ill adult neurointensive care patients during mechanical ventilation that mainstream ETCO2 did not provide a stable reflection of PaCO2 in all neurointensive care patients. ABG cannot be discarded when monitoring respiratory acid-base balance in mechanically ventilated neurointensive care patients.
Moreover, in contrary to our findings, in a prospective observational study, conducted on 180 intubated patients with trauma, the authors found a poor correlation between PaCO2 and ETCO2 (R2: 0.277), recommending nonuse of ETCO2 as a noninvasive method to reflect PaCO2, possibly owing to inability to record concurrent values of ETCO2 and PaCO2 for all the studied patients, as was declared in their study limitations [14].
In a study conducted on 35 patients who were endotracheal intubated as they were prepared for general anesthesia, the authors revealed that mean ETCO2 (29.1±2.5 mmHg) was less than mean PaCO2 (35.1±3.8 mmHg), and mean difference between PaCO2 and mainstream ETCO2 was 6 mmHg and limits of agreement were 11.8–0.3 by Bland-Altman comparison, which were in line with our results [15].
In accordance to our results, in a study conducted on 60 patients (30 COPD patients and 30 non-COPD patients) on mechanical ventilation who underwent prolonged weaning, comparison among PaCO2, ETCO2 and transcutaneous PCO2 (PtcCO2) showed that the mean of ETCO2 (36.5±7.5) was less than the mean of PaCO2 (42.4±8.6). In addition, the mean difference between PaCO2 and ETCO2 was 5.9±5.3 mmHg (95% CI: 4.5–7.2 mmHg), and the limits of agreement were −16.2 and 4.5 by Bland-Altman test. Worthwhile, the researchers found that PtcCO2 was more accurate than ETCO2 as it is dependent on skin vascularization, recommending its use, and they explained the slightly lower measurements of ETCO2 in comparison with PtcCO2 owing to that half of the studied patients had COPD (30 patients) with possible ventilation/perfusion mismatch, which could be responsible for lower results of ETCO2 when compared with PtcCO2[16].
Our results were in line with Cinar et al. [7] who studied 162 nonintubated adult patients presented with acute dyspnea who required ABG analysis that was compared with mainstream ETCO2. They revealed a highly statistically significant positive correlation between mainstream ETCO2 and PaCO2 (r=0.911, P<0.001). However, in contrast to our results, they found that the mean of ETCO2 level (39.47±10.84) was more than the mean of PaCO2 level (38.95±12.27). Moreover, the mean difference between PaCO2 and mainstream ETCO2 was −0.5±5 mmHg (95% CI: ̶̶ 1.32 to 0.27) and the limits of agreement were −10.5 and +9.5 mmHg by the Bland-Altman comparison, which were not matched with our findings. Consistent with our findings, Delerme et al. [17], in a prospective study, conducted on 43 nonintubated patients, presented with acute respiratory distress and admitted in emergency department, found a positive correlation between PaC02 and microstream ETCO2 (R=0.82). However, in contrary to our results, they reported that the mean difference between the PaC02 and microstream ETCO2 levels was 8±10 mmHg and limits of agreement were −10 to +26 mmHg, by the Bland-Altman comparison, in spite of usage of microstream capnography, which is an updated version of sidestream capnography, but they concluded possible reasons for inaccurate results as microstream capnography requires active cooperation of the patients, which might be more difficult to be achieved with their participant older patients presented by respiratory failure.
In contrast to our results, Kartal et al. [18] reported in a prospective cross-sectional study on 118 nonintubated patients with COPD presented by acute exacerbations a difference between PaCO2 and sidestream ETCO2 measurements of 8.4±11.1 mmHg and the limits of agreements were −13.4 to 30.2 by Bland-Altman comparison. These contradictory results might be owing to their usage of sidestream capnography, which was developed initially for nonintubated patients, which has several technical limitations, such as increased dead space, risk of occlusion by water, and secretions and dilution of the aspirated air that may lead to lower results, all of which may contribute to these inconsistent results.
Our study had several limitations. It was conducted in a single center with one set of medical devices. It involved a small number of patients having only respiratory and cardiac diseases with possible influence of ventilation/perfusion mismatch on the results. Moreover, we did not study the difference between various types of capnography (mainstream, sidestream, and microstream) with their effect on the measured ETCO2.
ConclusionETCO2 measurement is an essential tool for assessing ventilation and gives a proper estimation of PaCO2 in critically ill patients undergoing mechanical ventilation. It provides further knowledge about respiratory and cardiovascular function. Obviously, ETCO2 measurement is safe, noninvasive, and limits the need for invasive and repeated ABG samples.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)