Noninvasive carbon dioxide monitoring in pediatric patients undergoing laparoscopic surgery: transcutaneous vs. end-tidal techniques

Arterial carbon dioxide pressure (PaCO₂) is one of the most critical indicators of a patient’s respiratory function. The gold standard for measuring PaCO₂ in clinical practice is arterial blood gas analysis (ABGA) [1]. The patient’s acid–base balance and electrolyte state can be determined using ABGA. However, arterial blood sampling is an invasive procedure with risks of bleeding, infection, thrombosis, and vascular and neurologic harm. Additionally, ABGA cannot be used to monitor the PaCO₂ level continuously [2].

End-tidal carbon dioxide partial pressure (PetCO₂) has become a routine monitoring item for patients undergoing general anesthesia with tracheal intubation. Anesthesiologists can estimate PaCO₂ based on PetCO₂. However, several factors, including patient’s age, different types of surgery, combined cardiopulmonary diseases, and changes in pulmonary blood flow, can affect the accuracy of PetCO₂ monitoring results, increasing the difference between PetCO₂ and PaCO₂ in practice [3]. The correlation between PetCO₂ and PaCO₂ shows a decrease with the delay of pneumoperitoneum during laparoscopic surgery; thus, PaCO₂ values should be monitored intermittently by ABGA [4]. Therefore, PetCO₂ is not a reliable predictor of PaCO₂.

Transcutaneous carbon dioxide partial pressure (PtcCO₂) can be used to estimate PaCO₂. PtcCO₂ monitoring is based on an electrochemical principle. The probe’s internal heating electrode raises the local skin temperature. This results in the arterialization of dermal capillaries and improvement in their permeability, making it easier for CO₂ to enter the tissue space and diffuse away from the skin surface. CO₂ permeates the electrolyte layer via a high-permeability membrane on the sensor’s surface, altering the pH value of the electrolyte layer, which is related to the variation in the CO₂ partial pressure [5]. The PtcCO₂ value is obtained by the monitor’s internal programming algorithm. PtcCO₂ monitoring is a continuous and noninvasive method; however, due to advancements in monitoring technology and the miniaturization of the device, it is gaining increasingly popularity in clinical practice. Also, previous studies have shown that PtcCO₂ monitoring is effective in perioperative settings [6, 7], which suggests that compared to PetCO₂, PtcCO₂ has a better correlation and a smaller difference with PaCO₂. Nevertheless, Bolliger et al. indicated that PtcCO₂ monitoring does not accurately reflect PaCO₂ and does not provide more useful monitoring data than PetCO₂ [8].

Currently, there are no clinical reports regarding PtcCO₂ monitoring in pediatric laparoscopic surgery. This study aimed to investigate the correlation and consistency between PtcCO₂, PetCO₂, and PaCO₂ in children who underwent laparoscopic surgery (pneumoperitoneum time > 90 min).

The present study was approved by the Shenzhen Children’s Hospital Ethics Committee (Shenzhen, China; Ethics approval number: 202007402), and written informed consent was obtained from the parents. A total of 35 children who underwent laparoscopic surgery, aged 2–8 years, and ASA class Ι or II were recruited for this study. Patients who required a vasoconstrictor to maintain blood pressure during the procedure and those with insufficient pneumoperitoneum time (< 90 min) were excluded from this study.

Children were escorted to the anesthesia room by their parents, and 2.5 mg/kg propofol was administered to induce sleep. The child was then transported to the operating room under the close supervision of the anesthesiologist. Electrocardiogram, pulse oxygen saturation (SpO₂), and blood pressure (BP)were monitored; the heart rate (HR) and BP were recorded as baseline values. Tracheal intubation was conducted following intravenous administration of benzenesulfonate cisatracurium 0.1 mg/kg and fentanyl 3 μg/kg. Breathing settings were established on the anesthesia machine: intermittent positive pressure ventilation, tidal volume 6–10 mL/kg, inspiration/expiration 1/2, inhalation oxygen concentration 50%, gas flow rate 2 L/min, and respiratory rate 15–25 times/min. PetCO₂ level was continuously measured using the side-stream capnography (Datex-Ohmeda, Finland, air pumping speed 150 mL/min). The respiratory rate and tidal volume were adjusted to maintain PetCO₂ 35–45 mmHg and airway pressure 10–25 cmH₂O. Anesthesia was maintained with inhalation sevoflurane concentration at 2–3%, intravenous pumping of remifentanil 0.2 μg/kg/min, and benzenesulfonate cisatracurium 0.1 mg/kg/h. The dosage of benzenesulfonate cisatracurium was adjusted according to the results of muscle relaxation monitoring.

After tracheal intubation, a PtcCO₂ monitor (SenTec Digital Monitor, SenTec Inc. Therwil, Switzerland) was attached. The monitoring site was located on the forehead, and the electrode-heating temperature was adjusted to 42 °C. The monitor was calibrated, and the electrode membrane was changed before each use. Radial artery catheterization was performed to facilitate invasive arterial blood pressure monitoring and the acquisition of blood gas analysis samples. The laparoscopic pneumoperitoneum pressure was chosen based on the children's ages (2–4 years old: 9 mmHg, 5–8 years old: 11 mmHg) and it was fine-tuned according to the surgical field's size when the pneumoperitoneum has just been established. Throughout the course of the procedure, there was no change in the pneumoperitoneum pressure. HR and BP were maintained within the range of ± 20% of the baseline value. The nasopharyngeal temperature of the child was maintained at 36–37 °C using an inflatable thermal blanket; the operating room temperature was set at 23–25 °C.

PtcCO₂ monitor sensor was removed during the postoperative recovery period of anesthesia; the local skin was cleaned and examined for signs of injury. After the removal of the arterial cannula, pressure dressings were applied. The tracheal tube was withdrawn when the child’s spontaneous breathing was recovered, with SpO₂ maintained at 95% under inhaled air settings. Finally, the vital signs were observed carefully, and the child was transferred to the anesthesia recovery room for further monitoring after stabilization.

ABGA was conducted before (T0) and 30 min (T1), 60 min (T2), and 90 min (T3) after pneumoperitoneum. Also, PtcCO₂, PetCO₂, and PaCO₂ values were recorded at each time point. A blood gas analyzer (i-STAT Analyzer MN: 300-G, Singapore) was used to measure PaCO₂. HR, BP, SpO₂, tidal volume, respiratory rate, and body temperature at each time point. HR, BP, and the anesthesia machine’s respiratory parameters were stabilized for at least 5 min before recording the measured values.

Data were analyzed using SPSS version 26.0. Measurement data were presented as mean ± standard deviation (SD). Pearson’s correlation coefficient and linear regression analysis were conducted to establish the correlation, and the Bland–Altman method was utilized to assess the agreement between PetCO₂ and PaCO₂ or between PtcCO₂ and PaCO₂. A difference of ≤ 5 mmHg between PaCO₂ and the other two noninvasive variables was clinically acceptable [7, 9] and compared using the chi-square (χ²) test. P < 0.05 indicated a statistically significant difference.

Three children were excluded from the research due to the pneumoperitoneum’s short duration (< 90 min), and 32 children (21 men and 11 women; mean age 4.31 ± 1.65 years) were eventually recruited. The average height and weight were 103.63 ± 11.98 cm and 17.45 ± 3.75 kg. Among them, 26 patients underwent laparoscopic pyeloplasty, and 6 patients underwent laparoscopic choledochal cyst excision. A total of 128 datasets were collected, consisting of simultaneous measurements of PtcCO₂, PetCO₂, and PaCO₂ at four time points (Table 1). During the course of the study, no adverse events were observed.

The correlation coefficient r between PtcCO₂ and PaCO₂ was 0.84 (P < 0.01, n = 128), and the linear regression equation was PtcCO₂ = 7.89 + 0.82 × PaCO₂ (r² = 0.70, P < 0.01) (Fig. 1). A close correlation was established between PtcCO₂ and PaCO₂ at different monitoring time points; however, the correlation decreased slightly with increasing duration of pneumoperitoneum (Table 2).

Correlation coefficient r between PetCO₂ and PaCO₂ was 0.83 (P < 0.01, n = 128), and the linear regression equation was PetCO₂ = 9.87 + 0.64 × PaCO₂ (r² = 0.69, P < 0.01) (Fig. 2). A good correlation was established between PetCO₂ and PaCO₂ at various time points; however, as the pneumoperitoneum time extended, this correlation decreased significantly compared to PtcCO₂. The correlation was lowest at T3 (Table 3).

Difference between PtcCO₂ and PaCO₂ was 0.66 ± 2.51 mmHg (n = 128). The 95% limits of agreement (LOA) of PtcCO₂ – PaCO₂ average was 0.66 ± 4.92 mmHg (mean ± 1.96 SD) (Fig. 3). Difference between PetCO₂ and PaCO₂ was –4.4 ± 2.48 mmHg (n = 128). The 95% LOA of PetCO₂ – PaCO₂ average was –4.4 ± 4.86 mmHg (mean ± 1.96 SD) (Fig. 4). PtcCO₂ – PaCO₂ and PetCO₂ – PaCO₂ values at each time point are shown in Table 4. A difference of ≤ 5 mmHg was observed between PtcCO₂ and PaCO₂ in 122/128 samples and between PetCO₂ and PaCO₂ in 81/128 samples (P < 0.01).

Since the development of minimally invasive surgery, the laparoscopic operation has gained increasingly popular in pediatric surgeries due to its advantages of less trauma, a short hospital stay, less postoperative wound pain, and fewer complications. CO₂ is the most common gas utilized to create a pneumoperitoneum and provide a good operating view for the surgeon. However, the diffusion capacity of CO₂ is strong, and the absorption of CO₂ is sufficient in children due to factors such as the small volume of the abdominal cavity, the proximity of the capillaries to the peritoneum, and the larger abdominal surface area related to weight compared to adults [10]. A risk of hypercapnia is associated with prolonged artificial pneumoperitoneum. Increased CO₂ alters the body’s acid–base balance and stimulates sympathetic nerves, thus increasing catecholamine and cortisol release and leading to hemodynamic fluctuations [10, 11]. Close monitoring of the CO₂ level during laparoscopic surgery and timely adjustment of the ventilator parameters is essential to avoid the disruption of physiological functions. PaCO₂ levels are stabilized after 60 min of pneumoperitoneum [12]; hence, a pneumoperitoneum time of at least 90 min was appropriate to observe the variables in this investigation.

PetCO₂ is a routine measurement during the perioperative period and one of the primary indicators used to adjust ventilator parameters. However, factors that affect lung ventilation/perfusion may interfere with the accuracy of PetCO₂ measurements, and thus, the use of PetCO₂ in non-tracheal intubated patients is restricted. The increased abdominal pressure during laparoscopic surgery results in a diaphragmatic rise and an increase in thoracic pressure; subsequently, airway resistance and airway pressure also rise, with pulmonary vasoconstriction and reduced pulmonary blood flow. Pediatric patients are vulnerable to pneumoperitoneal pressure effects. This study revealed that during the entire monitoring process, a good correlation was established between PetCO₂ and PaCO₂, r = 0.83 (P < 0.01). Nevertheless, as the pneumoperitoneum time was prolonged, the correlation between PetCO₂ and PaCO₂ decreased gradually, which was consistent with previous findings [6, 9].

Several clinical studies have focused on the application of PtcCO₂ monitoring in different types of surgery and patients [13‐17] under non-tracheal intubation monitoring anesthesia [18, 19]. These studies confirmed the effectiveness of PtcCO₂ monitoring. The current results showed a close correlation between PtcCO₂ and PaCO₂, r = 0.84 (P < 0.01), and although the correlation was decreased with prolonged pneumoperitoneum time, it was not very significant compared to PetCO₂. According to Bland–Altman analysis, a lesser mean difference was detected between PetCO₂ and PaCO₂ than between PetCO₂ and PaCO₂. Therefore, PtcCO₂ performed better than PetCO₂ in estimating PaCO₂, which is in agreement with the previous results [6, 7, 9]. In our experiment, we can combine PetCO₂, PaCO₂, and PtcCO₂ to regulate the patient's acid–base, so there is no accumulation of CO₂ during the whole operation. However, due to the limitations of PetCO₂ monitoring, especially in the case of prolonged pneumoperitoneum, relying solely on PetCO₂ to regulate the patient's respiratory parameters cannot guarantee that the patient is in acid–base balance, particularly for young children. And PtcCO₂ can more accurately estimate PaCO₂, so its application can reduce the risk of CO₂ accumulation. Conway et al. conducted a meta-analysis on the effectiveness of PtcCO₂ monitoring [20] and demonstrated that is challenging to achieve a uniform standard due to the involvement of various clinical aspects, including the monitoring site, electrode heating temperature, and application population; thus, it is critical to monitor the PtcCO₂ trend throughout the monitoring process.

The CO₂ level measured by PtcCO₂ monitoring consists of two parts: one derived from the blood (arterial, capillary, and venous), and the other from the metabolism of the tissue cells [21, 22]. The warming effect of the electrode increases the skin blood flow and enhances the contribution of arterial blood to CO₂ by opening the precapillary sphincter [23]. A rise in the local skin temperature increases the metabolism of tissue cells, producing excessive CO₂. Therefore, the PtcCO₂ monitoring value is theoretically higher than that of PaCO₂. PtcCO₂ monitors used in clinical practice correct the initial measured value based on the selected heating temperature to reduce the deviation from PaCO₂ [21]. In the present study, PtcCO₂ monitoring values were less than PaCO₂ in 51/128 data sets; hence, the correction method for PtcCO₂ monitors needs to be investigated further.

Several factors affect PtcCO₂ monitoring, including the temperature of the electrodes, the monitoring location of the sensor, and the patient’s clinical state. Nishiyama et al. demonstrated that when the anterior chest (between the clavicle and nipple) was chosen as the monitoring site, PtcCO₂ correlated best with PaCO₂ at 43 °C (R² = 0.7568) among the different electrode-heating temperatures (37, 40, 42, 43, and 44 °C) in its setting, and the monitoring required less time to stabilize at higher temperatures as blood CO₂ levels change, but required > 150 s [24]. According to a study on the optimal electrode temperature for monitoring PtcCO₂ in preterm infants, the mean difference between PtcCO₂ and PaCO₂ was the smallest at 42 °C [25]. A higher temperature may result in skin damage in pediatric patients due to thin skin, but previous studies have not reported any skin injuries in children or infants. In this study, we chose 42 °C as the electrode temperature for PtcCO₂ monitoring; no adverse events were observed.

Nishiyama et al. reported that PtcCO₂ was correlated with PaCO₂ when the monitoring sensor was located on the chest (R² = 0.76) but not when it was located on the upper arm and forearm (R² < 0.5) [26]. When the anterior chest is chosen as a monitoring site in pediatric patients, the area of surgical disinfection might be affected, especially in younger kids. Anesthesiologists were usually positioned on the cephalic side of the patient, such that the forehead was selected as the site in this study, facilitating the administration of the probe. In the current study, PetCO₂ showed a close correlation with PaCO₂ than PtcCO₂ before pneumoperitoneum; however, the mean difference between PtcCO₂ and PaCO₂ was smaller than the mean difference between PetCO₂ and PaCO₂. However, whether PetCO₂ correlates better with PaCO₂ than PtcCO₂ in pediatric patients under non-pneumoperitoneal conditions with the forehead as the monitoring site needs to be studied further with a large sample size.

In the event that patients’ peripheral tissues and organs are not supplied adequately with blood, such as in shock, the CO₂ produced by tissue metabolism cannot be carried away quickly, and PtcCO₂ monitoring values increase gradually [22]. Thus, PtcCO₂ can be utilized as one of the indicators for assessing a patient’s microcirculatory status, which is useful in guiding the treatment [27]. However, the study on PtcCO₂ monitoring in surgical patients with circulatory failure has been rarely reported, and the correlation between the PtcCO₂ gradient changes and skin tissue perfusion status requires further clinical investigation. Other factors, such as poor skin contact with the fixed connection loop and insufficient gel, may allow contact between the probe and air, thus interfering with the monitoring results. CO₂ permeability films that have not been replaced for a long time or are damaged or air bubbles under the film can also affect the accuracy of PtcCO₂ monitoring.

Since PtcCO₂ monitoring is a continuous and noninvasive method that can be used to assess PaCO₂ to some extent, its perioperative application is promoted in the different types of surgery and populations. Endotracheal intubation is not required for gastrointestinal endoscopy or other operations that can be performed using nerve blocks. The use of intravenous anesthetic medications intraoperatively can improve operating conditions and increase patient comfort during these procedures. When the nerve block is unsatisfactory, or when specific operations call for an enhanced level of sedation, supplemental narcotic medicines are required. Understanding the CO₂ level of patients allows us to more precisely regulate the intravenous anesthetic medicine dosage. However, it is often difficult to accurately monitor the CO₂ levels of patients during these operations. In this situation, PtcCO₂ monitoring is a good choice. It has been shown [18, 19, 28] that PtcCO₂ monitoring is an effective way to detect hypoventilation in patients, which reduces the incidence, extent, and duration of hypercapnia, improving the safety of patients under sedation. High-frequency ventilation is often used to maintain oxygenation in some airway procedures performed with a rigid bronchoscope. However, evaluating the ventilatory status of patients with PetCO₂ in the open ventilation mode of high-frequency ventilation is challenging. As a result, we can adjust the parameters of high-frequency ventilation to avoid the accumulation of CO₂ according to PtcCO₂ [29].

Usually, patients undergoing thoracic surgery have chronic lung diseases and require one-lung breathing in the lateral decubitus position during operation. These factors can affect lung ventilation/perfusion, leading to the increase of the difference between PetCO₂ and PaCO₂, and patients are more likely to develop respiratory acidosis. Oshibuchi M's study [30] showed that PtcCO₂ can more accurately predict PaCO₂ compared with PetCO₂ in both two-lung ventilation and one-lung ventilation. It has been shown [31] that PtcCO₂ monitoring remains highly accurate even when one-lung ventilation is prolonged (more than 2 h) and permissive hypercapnia is present.

Most patients need to recover in anesthesia recovery room after surgery. Due to the presence of residual opioids and muscle relaxants, patients are at potential risk of developing respiratory depression, especially in elderly and obese patients. PtcCO₂ monitoring effectively reflects PaCO₂ levels and is more suitable for observing changes in CO₂ fluctuations over time so that we can take appropriate treatment measures [32]. For pediatric patients, the anesthesia has certain particularity. Children are often unable to cooperate with us for some examination operations. They must be under sedation and analgesia conditions prior to nerve block or spinal anesthesia. Children must maintain a certain level of sedation throughout the whole operation, but the use of anesthetic drugs will always have an impact on their breathing more or less. By using PtcCO₂ monitoring, anesthesiologists are able to determine in time whether CO₂ accumulation in patients is occurring so that the appropriate treatment can be administered.

PtcCO₂ monitoring also has some limitations. The monitoring site should be cleaned in advance to remove the hair and grease; also, the PtcCO₂ monitor requires a calibration time of approximately 15 min before use and needs to be recalibrated either after the patient is removed from the monitoring site or after prolonged monitoring. When CO₂ in the blood changes, PtcCO₂ monitoring takes about 2 min to reflect PaCO₂ with a degree of delay [33]. These factors limit its use in surgery patients. Therefore, the PtcCO₂ monitor needs further improvement to facilitate its use during the perioperative period. Some studies have reported that PtcCO₂ monitoring techniques are not based on electrochemical principles [34‐36]. Because of the different monitoring mechanisms, the local heating on the skin is avoided, and the time required for calibration and stabilization is short, rendering the monitors convenient to use. However, the application is still not mature in clinical practice. Although PetCO₂ is susceptible to various factors, it plays an essential role in determining the position of the tracheal tube, tube folding, and accidental decannulation [37]. Additionally, PetCO₂ provides information about the patient’s pulmonary blood flow status and circulatory function [38], and the patient’s airway status can also be determined from the PetCO₂ waveform, thereby deeming that PtcCO₂ is not a substitute for PetCO₂.

In conclusion, PtcCO₂ shows a close correlation with PaCO₂ when the forehead is chosen as a monitoring site in children undergoing laparoscopic surgery. Compared to PetCO₂, PtcCO₂ can accurately estimate PaCO₂ and could be used as an auxiliary monitoring indicator to optimize anesthesia management for laparoscopic surgery in children; however, it is not a substitute for PetCO₂.