Abstract

Background

High-concentration oxygen delivery via a face mask (FM) with a reservoir bag is a common practice to prevent postoperative hypoxemia; however, it may also lead to atelectasis and other respiratory complications. Lower concentrations delivered via nasal cannula (NC) may be equally effective in preventing postoperative hypoxemia. The present study aimed to compare peripheral oxygen saturation (SpO₂) delivered via NC versus FM with a reservoir bag in patients who have undergone general anesthesia (GA).

Methods

Eighty-four patients scheduled for GA were randomized to receive either oxygen via NC (NC group, n = 42) or FM with a reservoir bag (FM group, n = 42) for 30 minutes after GA at a postanesthesia care unit (PACU). All patients were assessed based on SpO₂ value, adverse events, and patient satisfaction (measured using a 100-mm visual analog scale).

Results

The overall difference between groups in the change of SpO₂ over 30 minutes at the PACU was –0.004 (95% confidence interval, –0.015 to 0.008; P = 0.527). SpO₂ during the first five minutes was lower in NC group, but the difference was not statistically significant. No desaturation occurred in either group, and there was no observed difference between groups in terms of adverse events. Patient satisfaction scores were also similar (P = 0.612).

Conclusions

Oxygen supplementation via NC and via FM with a reservoir bag were equally effective in preventing postoperative hypoxemia after GA.

Keywords

face mask with reservoir bag, general anesthesia, nasal cannula, oxygen saturation, postoperative hypoxemia

---

Introduction

Hypoxemia is a common complication in the early postoperative period after general anesthesia (GA)^1,2 due to incomplete lung re-expansion or the residual effects of anesthetic drugs leading to atelectasis, ventilation–perfusion mismatch, alveolar hypoventilation, or impaired upper airway patency.^3,4 Hypoxemia can cause pathophysiological disorders, brain damage, and an increased risk of mortality.⁵ Additionally, hypoxemia results in unplanned admission to an intensive care unit and a prolonged hospital stay.^6,7

An association has been reported between postoperative hypoxemia on arrival at the postanesthesia care unit (PACU) and not receiving oxygen supplementation during patient transfer.⁸ Oxygen supplementation has thus been recommended for patients recovering from GA in the immediate postoperative period, especially during transfer to the PACU.⁹ Various oxygen delivery devices are available, including a face mask (FM) with or without a reservoir bag and nasal cannula (NC).¹⁰ In our practice, patients recovering from GA receive oxygen using an FM with a reservoir bag at a flow rate of 8–10 L/minute, which provides oxygen at a high concentration during transfer to the PACU.

However, increasing the concentration of oxygen does not necessarily lead to better prevention of hypoxia after GA.¹¹ In fact, high oxygen concentrations can cause absorptive atelectasis and counteract the effects of hypoxic pulmonary vasoconstriction. Additionally, it can impair the detection of hypoventilation, reduce cardiac output, and increase systemic vascular resistance.⁴ Some studies have shown that postoperative hypoxia is uncommon and easily manageable, and the majority of patients in the PACU received oxygen supplementation without any specific indication.¹²

Furthermore, providing oxygen through an FM has several disadvantages. Often, because the mask may not fit the patient’s face, this reason causes a patient to become uncomfortable, and if the patient either removes the device or places it in an inappropriate position, this will result in clinical hypoxemia.¹³ Additionally, removal of the FM to suction secretion or assess the patient’s speech after surgery can cause temporary hypoxia.⁵ Some studies have reported that patients feel more comfortable with an NC, and NC also helps to reduce rebreathing of carbon dioxide.¹⁴

In our center, routine postoperative oxygen supplementation was given by FM with a reservoir bag at a flow rate of 8–10 L/minute, carrying approximately 0.7–0.9 fractional inspiratory oxygen (FiO₂) value. We hypothesize that given O₂ via NC at a flow rate of 4 L/min, the lower FiO₂ (0.36) would be effective in preventing hypoxemia.

This study aimed to compare the efficacy of an NC versus FM with a reservoir bag for preventing hypoxemia in patients recovering from GA at the PACU.

Methods

Design and Participants

This prospective randomized controlled trial was conducted at Khon Kaen University’s Srinagarind Hospital.The study protocol was approved by the Khon Kaen University Ethics Committee for Human Research in accordance with the declaration of Helsinki and ICH good clinical guidelines (HE611246; Thai Clinical Trials Registry identification number: TCTR20210824001). Data were collected between July 2018 and January 2019.

Eligible patients were those who underwent elective surgery under GA with an endotracheal tube (ETT), were between 18–60 years of age, and with the American Society of Anesthesiologists physical status (ASA PS) classification of 1–3. The exclusion criteria were obstructive sleep apnea, chronic obstructive pulmonary disease, emphysematous lung, severe pulmonary disease, heart disease, body mass index (BMI) > 35 kg/m², predicted difficult airway, baseline oxygen saturation (SpO₂) less than 94%, chronic use of oxygen therapy, airway or trans-nasal surgery, and any condition/factor that can interfere with pulse oximetry (peripheral artery disease, scleroderma, anemia, body temperature less than 34°C, hemodynamic instability, or use of intraoperative dye).

Procedure

Patients scheduled for elective surgery under GA were approached about participating in the study on the evening before their operation. After informed consent was obtained, participants were randomly assigned to either the NC or FM with a reservoir bag group. Randomization was carried out using a computer-generated block-of-four list (http://www.randomizer.org/). The NC group received an oxygen flow of 4 L/minute, and the FM group received an oxygen flow of 8 L/minute. The allocation sequence was sealed in opaque envelopes, which were opened at the end of the operation.

Surgery was performed using standard monitors under GA with an ETT. The SpO₂ level was recorded at baseline prior to preoxygenation. The anesthesia was maintained with a 2% expired sevoflurane concentration and a combination of oxygen and air to maintain a 50% oxygen concentration. The ventilator was set to produce a tidal volume of 6–8 mL/kg of predicted body weight, and then the respiratory rate was adjusted to maintain an exhaled carbon dioxide (EtCO₂) level between 30 and 40 mmHg. Thirty minutes before the end of the surgery, 8 mg of intravenous ondansetron was administered to prevent postoperative nausea and vomiting. At the end of the surgery, 0.02 mg/kg of atropine and 0.05 mg/kg of neostigmine were administered as reversal agents. Sevoflurane was discontinued and 100% oxygen was administered at a flow rate of 6 L/minute. The extubation criteria included hemodynamic stability, spontaneous breathing with tidal volume greater than 5 mL/kg, SpO₂ greater than 94%, EtCO₂ less than 50 mmHg, intact gag reflex, following commands, sustained head raise for 5 seconds, and hand grasp. Upon meeting the extubation criteria the ETT was removed, and the patient was provided with the oxygen delivery device determined by their randomized group allocation. Patients were then transferred to the PACU in a 20°–30° head-up position. Patients from whom the ETT could not be safely removed would be withdrawn from the study.

The patients received standard postoperative care at the PACU by nurse anesthetists. All participants continued receiving oxygen supplementation for 30 minutes, at which point they were weaned off of their oxygen delivery device. Patients who met PACU discharge criteria and who were able to maintain an SpO₂ level of at least 92% after device removal were transferred to the ward. If a patient was unable to wean off the oxygen, oxygen supplementation would continue, and the anesthesiologist would be notified so that they could investigate the cause and provide appropriate treatment.

Data Collection

Baseline characteristics included age, weight, height, BMI, ASA PS classification, underlying diseases, history of smoking, duration of anesthesia, and type of surgery.

SpO₂ was measured at 0, 5, 10, 15, and 30 minutes after arrival in the PACU (T0, T5, T10, T15, and T30, respectively) by nurse anesthetists. The number of patients whose devices were removed for reasons such as discomfort or to suction secretions was recorded, as were cases of breathing difficulty (respiratory rate more than 25 times per minute, accessory respiratory muscles breathing pattern, abnormal breathing sounds such as wheezing or stridor, cyanosis) and nausea/vomiting. Patient satisfaction with the oxygen delivery device was assessed using a visual analog scale (VAS) from 0 to 10 at the time of discharge from the PACU.

Outcome Measures

The primary outcome was the difference in change of SpO₂ within 30 minutes of arrival at the PACU following GA. Secondary outcomes were the incidence of desaturation (SpO₂ levels less than 94%), discomfort caused by the oxygen delivery device, breathing difficulty, nausea and vomiting, and oxygen device removal, as well as patient satisfaction.

Sample Size Calculation

The sample size was calculated based on the hypothesis that the two modes of oxygen delivery would be similarly effective in maintaining postoperative SpO₂. The required sample size was calculated based on an estimated greatest variance of 3.71 and a mean difference between groups of 1.61, which were derived from prospective randomized controlled trials and repeated measurement studies.⁵ We chose a beta error of 0.20 (80% power) and an alpha error of 0.05. According to these calculations and assuming a drop-out rate of 10%, 42 participants per group were required.

Statistical Analysis

All data were analyzed using STATA for Windows version 16.0 (StataCorp LP, College Station, TX, USA). Categorical data were presented as number and percentage (%), and continuous data were presented as mean and standard deviation (SD) or the median and interquartile range. All variables were tested for normal data distribution. An independent sample t-test or Mann–Whitney U test was used to evaluate continuous data, and a Chi-square test or Fisher’s exact test was used for categorical data. A generalized estimating equations (GEE) model was used to compare intergroup differences in changes in SpO₂ at T0–T30. A P value less than 0.05 was considered statistically significant.

Results

All of the 84 patients enrolled (42 in each group) completed the study (Figure 1). There were no major differences in baseline characteristics or intraoperative data (gender, age, body weight, BMI, ASA PS classification, comorbidities, history of smoking, duration of anesthesia, and type of surgery) between groups (Table 1). Most (66.7%) were female, and the average age was 41.8 years. Nearly all patients (98.8%) underwent surgery of areas other than the upper abdomen. Only one patient in the NC group (2.4%), who underwent upper abdominal surgery, had no adverse outcomes observed.

Mean SpO₂ values during the first 30 minutes in the PACU are shown in Table 2. Preanesthesia SpO₂ was 99.6 ± 0.9 in the NC group and 99.4 ± 0.9 in the FM group (P = 0.298). In the FM group, SpO₂ after PACU admission was 100 at all-time points. In NC group, SpO₂ at arrival (T0) and at 5 minutes after PACU admission (T5) were non-significantly lower than in the FM group but 100 at all later time points (T10, T15, T30). The overall difference in SpO₂ change throughout the course of study was –0.004 (95% confidence interval, –0.015 to 0.008; P = 0.527) according to GEE (Figure 2). At T0 point, there were four patients in the NC group with SpO₂ levels below 100% (three with 99% and one with 97%), leading to a relatively elevated SD.

Table 3 shows patient satisfaction with their oxygen delivery device and related adverse events. Both groups had high satisfaction scores (VAS greater than 9/10), with no significant difference (P = 0.612). No participants in either group experienced desaturation. In the FM group, three patients (7.1%) reported device discomfort versus just one patient (2.4%) in the NC group. However, this difference was not statistically significant (P = 0.616), nor were those in the incidence of breathing difficulty or nausea/vomiting. Oxygen delivery devices were removed from more patients in the FM group, but this difference was also non-significant (P = 0.359). Two patients had their FM removed during PACU admission (4.8%): one who experienced breathing difficulty and the other to speak with the nurse. Neither patient, however, experienced desaturation. No devices were removed in the NC group.

Discussion

Hypoxemia may develop in patients who have undergone GA due to a combination of atelectasis, ventilation-perfusion mismatch, alveolar hypoventilation, and impaired upper airway patency. Patients recovering from GA may benefit from brief supplemental oxygen administration to avoid hypoxemia.⁴ The most common oxygen delivery method immediately following GA is the FM with a reservoir bag, which delivers oxygen at high concentrations.⁴

In this study, average SpO₂ in both groups was higher than 99%, and there was no difference between groups in SpO₂ change at the PACU. At T5, SpO₂ was lower in the NC than the FM group, likely due to the lower oxygen flow provided by the former (4 vs. 8 L/minute) resulting in a lower FiO₂ (approximately 0.36 vs. 0.70). However, this difference was not statistically significant, suggesting that both methods were equally effective in preventing hypoxemia in patients recovering from GA. This result is consistent with those reported by Nolan et al.¹⁵, who compared postoperative oxygen supplementation between a flow of 4 L/minute via FM and 2 L/minute via NC in patients undergoing abdominal surgery. They found that NC delivery resulted in adequate SpO₂ levels, albeit lower than those in the FM group.

However, this contrasts with the findings of Ayhan et al.⁵, who compared average SpO₂ in patients who received oxygen at a flow rate of 5 L/minute through either NC or FM for 30 minutes after thyroidectomy at the PACU. They found that the NC group had significantly higher SpO₂ than the FM group. This discrepancy may have resulted from the fact that their FM group had removed their devices more often (75.5% vs. 3.8%) and had significantly higher BMI (27.3 kg/m² vs. 25.3 kg/m²) than the NC group.⁵

Additionally, none of our participants experienced desaturation during the study, which aligns with the findings of Raksakietisak et al.¹⁶, who compared NC and FM at flow rates of 4 and 5 L/minute, respectively, and reported zero incidence of hypoxemia. Another observational study by Siddiqui et al.¹⁷ found the incidence of hypoxemia in the PACU to be 19.2% in patients transferred without oxygen supplementation, but only 0.8% in those transferred with supplementation. Postanesthetic oxygen desaturation is frequently observed during the transport period and after arrival at the PACU.⁷ Recently published studies have found transport without oxygen to be the most significant predictor of desaturation.^8,17 The absence of desaturation in our study was thus likely due to oxygen being administered in the operating room immediately after ETT removal and continuously during transport to the PACU. In addition, the incidence of oxygen device removal was minimal (2.4%) and the duration of removal was short, so it did not lead to hypoxia or affect average SpO₂.

In our study, only 2.4% (2/84) of FMs were removed, which was significantly lower than the removal rates in Ayhan et al.’s study⁵ described above. The difference may be due to the fact that 37.8% of patients in their FM group experienced nausea and vomiting compared to none in ours. This, along with patient discomfort and the need to suction oral secretions, lead to frequent removal of FMs. Their study was also in thyroidectomy patients, who require regular FM removal for speech quality evaluation.

Only one patient in the FM group experienced breathing difficulty, similar to the findings of Raksakietisak et al.¹⁶ This may have been due to the fact that we excluded patients who were at risk of respiratory complications such as those with obesity or chronic respiratory diseases. Additionally, few patients underwent upper abdominal surgery, which carries a higher risk of respiratory complications.

In our study, patients in both groups reported similarly high satisfaction on the VAS (> 9/10), which differs from the results of Ayhan et al.⁵, where patients in the NC group had higher satisfaction scores (9.1 vs. 6.7/10). Their FM group reported higher distress, nausea, vomiting, and the need to remove a device during a test speech than the NC group, which could be the cause of this difference.⁵

Opting for an NC over an FM with a reservoir bag may also help reduce hospital costs, as FMs cost 58 THB (1.5 USD)/piece compared to just 18 THB (0.5 USD)/piece for NCs. As approximately 15,000 patients receive GA at our center each year, switching to NC could lead to an annual cost reduction of up to 540,000 THB.

We recommend initiating oxygen supplementation in GA patients as soon as they are extubated and then continue to provide oxygen throughout the transfer to the PACU. We found that supplemental oxygen delivered via NC and FM with a reservoir bag to be equally effective, in patients at low risk of hypoxemia.

One limitation of our study was that all of our participants were at low risk of hypoxemia, which means that our results may not be generalizable to higher-risk populations such as those with obesity or chronic respiratory disease or those undergoing upper abdominal or thoracic surgery. Moreover, we conducted this study in only a single hospital. Further multicenter studies that include patients at high risk of hypoxemia are therefore necessary to confirm our results.

Conclusion

Oxygen supplementation via NC is non-inferior to FM with a reservoir bag in preventing hypoxemia in patients who have undergone GA. There were no significant differences in terms of device-related complications or patient satisfaction.

Acknowledgments

This study was supported by the Khon Kaen University Faculty of Medicine Research Affairs unit (Thailand; grant number IN62104). We would like to acknowledge Dr. Dylan Southard, for editing this manuscript via the KKU Publication Clinic.

Conflict of Interest

The authors declare that there are no financial or other conflicts of interest.

---

References