High-flow nasal cannula oxygenation in comparison with apnoeic oxygenation during foreign body removal by rigid bronchoscopy: A randomised controlled trial.

Background and Aims: This trial aimed to assess the safety and efficacy of oxygenation using a high-flow nasal cannula (HFNC) compared with the standard apnoeic oxygenation technique during foreign body (FB) removal by a rigid bronchoscope.
Methods: A prospective, blinded, randomised, controlled trial was conducted on subjects planned for FB removal by a rigid bronchoscope. Inclusion criteria were male and female patients aged between 10 and 40 years. The primary outcome was the lowest oxygen saturation level monitored by pulse oximeter during the procedure, and the secondary outcome was the incidence of postoperative atelectasis.
Results: Nearly 64 patients were randomised into either HFNC oxygenation (N = 32) or apnoeic oxygenation (N = 32) group. There were no significant differences between the studied groups regarding age, gender, weight, or cardiovascular variables. Intraoperative oxygen saturation was 4% higher in the HFNC group at different times. The end-tidal carbon dioxide measured on induction and after termination by 15 min showed no significant differences between the two groups; however, it was significantly higher in the apnoeic oxygenation group immediately after the procedure, after 5 min, and after 10 min of termination. No adverse effects from the increased end-tidal carbon dioxide were reported.
Conclusion: HFNC was superior to apnoeic oxygenation technique in maintaining oxygenation status in patients undergoing rigid bronchoscopy for FB removal. Copyright:

INTRODUCTION

Bronchoscopy commonly produces hypoxaemia. The partial pressure of oxygen (PaO2) usually decreases about 20 mmHg during the procedure, and the worst decrease occurs during bronchoalveolar lavage. To avoid bronchoscopy-induced hypoxaemia, oxygen supply can be delivered by low or high gas flows. Recently, the high-flow nasal cannula (HFNC) has been introduced for oxygen therapy.[12] This delivers high flow, heated, humidified air via nasal prongs at a prescribed fraction of inspired oxygen (FiO2) and a maximum flow of 60 L/min. It maintains blood oxygenation during the apnoea period of bronchoscopy, generates a flow-dependent positive airway pressure, and improves oxygenation by increasing end-expiratory lung volume.[3] The aim of this trial was to assess the safety and efficacy of oxygenation using a HFNC compared with the standard apnoeic oxygenation technique during foreign body (FB) removal by a rigid bronchoscope.

METHODS

This trial was approved by our institutional Ethics Committee and was conducted in accordance with the Helsinki Declaration of 1975, as revised in 2013. The trial was registered with the clinical trial registry (ID: NCT04885673). We obtained informed written consent from the participants or their parents, and we were responsible for maintaining the confidentiality of the data.

This prospective, blinded, parallel-group (1:1 allocation ratio), randomised, controlled clinical trial was conducted at the University Hospitals (Cardiovascular Surgery Hospital, Thoracic Surgery Unit), between April 25, 2021 and July 5, 2021.

The effect size for the difference between proportions (Cohen’s effect) was calculated by the equation: h = 2−arcsin (√p1) – 2−arcsin (√p2), giving approximated h = 0.5, where p1 and p2 are the two proportions [(HFNC group (34.6%) and standard apnoeic oxygenation group (12.0%)] and using the G power programme (Version 3.1.9.4) for sample size calculation, setting power at 80%, alpha error 5%, and after reviewing previous study results[4] assuming medium effect size difference (0.4) in oxygenation levels [lowest peripheral oxygen saturation <90%] between two groups of patients planned for FB removal (one group had HFNC [34.6%] and the other had standard apnoeic oxygenation [12.0%]) and after considering dropout rate 20%; a sample size of at least 64 patients planned for FB removal (32 patients in each group) was sufficient to achieve the trial objective.

The sealed, opaque, sequentially-numbered envelopes method was used for randomisation and allocation concealment. We prepared two sets of 32 identical, opaque, letter-sized envelopes. Each envelope contained a white allocation paper, marked as “Treatment A” (n = 32) or “Treatment B” (n = 32) and a sheet of single-sided carbon paper closest to the front of the envelope (with the carbon side facing the white paper). Finally, we sealed the envelopes and signed across the seal. We combined the two sets (64 envelopes) and shuffled them thoroughly. Then, we marked a number on the front of each envelope sequentially from 1 to 64 and placed them into a plastic container in numerical order. An investigator (not involved in sequence generation and allocation concealment) assessed participants for eligibility and assigned eligible patients to receive either HFNC oxygenation (group I) or apnoeic oxygenation (group II). Participants, health assessors, and data analyst were blinded to treatment allocation.

We included male and female subjects (10–40 years old) with recent FB aspiration in the last 72 h (based on history or presence of a radiopaque shadow in their chest radiogram or computed tomography scan) who were American Society of Anesthesiologists physical status I (ASA PS I) and were scheduled for FB removal by a rigid bronchoscope. Scheduled or emergency cases were included. The exact position and type of FB were not assessed in this trial. Patients were randomly assigned into two groups. We excluded patients who were critically ill, intubated, or weighing <40 kg.

Preoperative investigations were done as per our institutional protocol. Complete blood count, coagulation profile, and chest radiogram were ordered for all patients prior to the planned procedure except in an emergency situation. For all patients included in this trial, preoperative assessment including complete airway evaluation (mouth opening, Mallampati grading, thyromental distance, and evaluation of dentition) was performed. Standard fasting guidelines were applied except on emergency hypoxic patients. An intravenous (IV) anti-aspiration prophylaxis (0.1 mg/kg of ondansetron) was administered to all patients. The risk of airway oedema after the procedure was avoided by preoperative administration of dexamethasone (0.2 mg/kg, IV), and it was treated by adrenaline nebulisation 1:1000 in 3 mL of normal saline.[5]

On the patient’s arrival to the operation room, an IV line was secured, and an IV dose of atropine (0.01 mg/kg) was administered. Standard monitoring, including non-invasive blood pressure (BP), electrocardiography (ECG), pulse oximetry, and capnography were applied for all patients. Capnography was performed in the early stages by mask ventilation, but the mask was removed at the time of insertion of the bronchoscope.

Anaesthesia was induced using propofol (2–3 mg/kg, IV) administered during mask preoxygenation with 100% oxygen, followed by a non-depolarising muscle relaxant (atracurium, 0.5 mg/kg) and fentanyl (0.5 mg/kg).[5]

After the achievement of anaesthesia and full muscle relaxation, ventilation was started by the open mask technique. Then, the patients were randomised into two groups according to the oxygenation technique used. In group A (N = 32), an HFNC Vapotherm (Precision Flow Plus, Vapotherm®, Exeter, New Hampshire, USA) was used during bronchoscopy time using 100% humidified oxygen, given at a flow rate of 50 L/min, and the temperature was adjusted at 38°C.[4] In group B (N = 32), the standard apnoeic oxygenation technique was applied with 100% humidified oxygen, administered at a flow rate of 6 L/min through the side port of the bronchoscope during the procedure.[6]

Next, the surgeon inserted the rigid bronchoscope through the glottis opening and examined the airway searching for the FB. Heart rate, blood pressure, and oxygen saturation were recorded every 5 min till the end of the procedure. If oxygen saturation dropped below 92%, the surgeon was informed to stop the procedure and to start mask ventilation till the improvement of oxygenation, with the capnogram attached to the mask to record the end-tidal carbon dioxide level. The time of the procedure ranged from 30 to 60 min.

During the procedure, anaesthesia was maintained by IV infusion of propofol (25–50 mg/kg/min) with incremental doses of atracurium (0.1 mg/kg, IV) every 20 min.[5]

After removal of the FB, the bronchoscope was withdrawn, and the HFNC was removed. Then, mask ventilation (with capnogram attached to the mask) was started till full recovery from muscle relaxation and regaining of consciousness. The patient was transferred to the postoperative (PO) intensive care unit (ICU), and his heart rate, blood pressure, and oxygen saturation were recorded. A chest radiogram was obtained 6 h after the procedure to detect basal atelectasis, if any.

The primary outcome was maintaining oxygen saturation above 92% during the procedure. Monitoring oxygen saturation was carried out using a pulse oximeter, and the readings were recorded.

Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) Statistics for Windows,version 22.0(International Business Machines Corp.,Chicago,USA). For quantitative data, the Shapiro–Wilk test for normality was performed. For data that followed a normal distribution, values were expressed as mean ± standard deviation (SD). Comparisons between two groups were carried out using the Independent Samples T-test. For qualitative data, the variables were summarised as frequencies. Pearson’s Chi-square test for independence or Fisher’s exact test was used to examine the association between two categorical variables as appropriate. A P value <0.05 was adopted to interpret the significance of statistical tests.

RESULTS

Around 91 subjects undergoing FB removal with a rigid bronchoscope were assessed, and 64 met eligibility criteria (all patients were scheduled, and there were no emergency cases) and were randomly allocated to receive oxygenation through either an HFNC (N = 32) or the standard apnoeic oxygenation technique (N = 32) [Figure 1].

Figure 1

The trial flow diagram

The two groups were comparable regarding age, gender, weight, and bronchospasm at presentation with no significant differences (P > 0.05) [Table 1].

Table 1

Demographic characteristics among the studied groups

Variables	HFNC group (n=32)	Apnoeic oxygenation - AO group (n=32)	P
Age (years), mean±SD	17.6±4.0	18.5±5.7	0.464
Gender (n, %)
Male	12 (37.5%)	13 (40.6%)	0.798
Female	20 (62.5%)	19 (59.4%)
Weight (kg), mean±SD	64.5±12.5	68.5±13.7	0.226
Bronchospasm (n, %)	13 (40.6%)	15 (46.9%)	0.614

*Significant. SD: Standard deviation

The need for post-procedural invasive ventilation was non-significantly less frequent in the HFNC group. The attempts number and the procedure duration were significantly lower in the HFNC group [Table 2].

Table 2

Operative findings among the studied groups

Variables	HFNC group (n=32)	AO group (n=32)	P	Effect size
Invasive ventilation (n, %)	2 (6.3%)	6 (18.8%)	0.26	Relative risk (95% CI) 0.33 (0.07-1.53)
Attempts number (mean±SD)	5.1±0.9	6.3±1.1	<0.001*	Mean±SE (95% CI) −1.1±0.3 (−1.6 to 0.)
Procedure duration (min), mean±SD	20.7±6.2	34.5±5.4	<0.001*	−1.1±0.3 (−1.6 to 0.)

*Significant; Effect size: value of HFNC relative to AO; CI: confidence interval; SE: standard error; SD: Standard deviation

We found no significant differences between the studied groups regarding the intraoperative (IO) temperature, heart rate, and mean blood pressure. IO oxygen saturation was higher in the HFNC group; the differences were statistically significant at different times except at the baseline and 5 min. The end-tidal carbon dioxide levels measured on induction and after termination by 15 min showed no significant differences between the two groups. Immediately after termination of the procedure, after 5 min, and after 10 min, the end-tidal carbon dioxide levels were significantly higher in the group of apnoeic oxygenation compared to the HFNC group. As regards the assessment of PO atelectasis by chest radiogram in the ICU, it was significantly less frequent in the HFNC group when compared to the preoperative one. There were no emergency cases in our trial [Table 3].

Table 3

Postoperative findings among the studied groups

Variables	Time	HFNC group (n=32)	AO group (n=32)	P	Effect size Mean±SE (95% CI)
Intraoperative
Temperature (℃)	Baseline	37.2±0.1	37.2±0.2	0.71	0.0±0.0 (−0.1 to 0.1)
	Minute-5	37.1±0.1	37.2±0.2	0.33	0.0±0.0 (−0.1 to 0.0)
	Minute-10	37.1±0.1	37.2±0.2	0.2	−0.1±0.0 (−0.1 to 0.0)
	Minute-15	37.1±0.1	37.1±0.2	0.21	−0.1±0.0 (−0.1 to 0.0)
	Termination	37.1±0.1	37.0±0.2	0.37	0.1±0.0 (0.0-0.1)
Heart rate (beats/min)	Baseline	93.2±2.9	92.9±2.8	0.73	0.3±0.7 (−1.2 to 1.7)
	Minute-5	101.2±3.4	101.1±3.0	0.94	0.1±0.8 (−1.5 to 1.6)
	Minute-10	99.2±3.1	99.1±3.0	0.9	0.1±0.8 (−1.4 to 1.6)
	Minute-15	97.2±3.3	96.9±3.1	0.76	0.3±0.8 (−1.3 to 1.8)
	Termination	94.7±3.3	93.9±3.0	0.33	0.8±0.8 (−0.8 to 2.4)
Mean blood pressure (mmHg)	Baseline	86.1±4.9	86.5±4.8	0.74	−0.4±1.2 (−2.8 to 2.0)
	Minute-5	93.0±5.0	93.6±4.9	0.67	−0.5±1.2 (−3.0 to 1.9)
	Minute-10	91.2±4.8	91.5±5.0	0.8	−0.3±1.2 (−2.8 to 2.1)
	Minute-15	89.4±4.8	89.6±4.8	0.86	−0.2±1.2 (−2.6 to 2.2)
	Termination	88.5±5.0	88.6±4.9	0.9	−0.2±1.2 (−2.6 to 2.3)
Oxygen saturation (%)	Baseline	97.9±0.8	97.9±0.8	0.74	0.1±0.2 (−0.3 to 0.4)
	Minute-5	97.4±0.8	97.0±0.9	0.08	0.4±0.2 (0.0-0.8)
	Minute-10	96.6±0.9	95.1±1.2	<0.001*	1.5±0.3 (0.9-2.0)
	Minute-15	96.1±1.1	94.6±1.1	<0.001*	1.5±0.3 (1.0-2.0)
	Termination	95.4±1.2	92.0±0.8	<0.001*	3.5±0.3 (3.0-4.0)
Intraoperative (IO) and postoperative (PO)
End-tidal carbon dioxide (mmHg)	Baseline (IO)	33.0±1.1	33.3±1.0	0.34	−0.3±0.3 (−0.8 to 0.3)
	Termination (IO)	51.8±1.7	55.2±2.7	<0.001*	−3.4±0.6 (−4.5 to 2.)
	Minute-5 (PO)	43.0±1.6	45.9±2.4	<0.001*	−3.0±0.5 (−4.0 to 2.)
	Minute-10 (PO)	39.1±1.5	40.2±1.5	0.003*	−1.2±0.4 (−1.9 to 0.)
	Minute-15 (PO)	35.9±1.5	36.4±0.7	0.06	−0.6±0.3 (−1.1 to 0.0)
Atelectasis, (n, %)		1 (3.1%)	8 (25.0%)	0.03*	Relative risk (95% CI) 0.13 (0.02-0.09)

SE: Standard error; CI: Confidence interval

DISCUSSION

In this trial, the primary outcome was maintaining oxygen saturation >92% during the procedure. Monitoring oxygen saturation was carried out using a pulse oximeter, the readings were recorded, and the secondary outcome was the incidence of PO atelectasis.

Gas exchange is usually impaired during bronchoscopy owing to sedation and ventilation–perfusion mismatching. Increased airway resistance due to the presence of the bronchoscope and gas aspiration through it may result in atelectasis.[7] HFNCs have recently been introduced for oxygen therapy, but they have not been used so far during bronchoscopy in paediatrics.[8] Therefore, the current trial aimed to assess the safety and efficacy of oxygenation using an HFNC in patients undergoing rigid bronchoscopy.

Usage of HFNC seems to improve alveolar recruitment, increase functional residual capacity, and increase intrathoracic pressure mostly because of positive end-expiratory pressure (PEEP) added by the HFNC. The definitive level of the provided PEEP by high-flow devices is still debatable, best estimates are 1 cm H2O of PEEP for every 10 L/min of flow delivered with closed mouth breathing. This may vary from one patient to another as there are many factors that can affect the PEEP level delivered, such as the patient’s size and age.[9]

The heated, humidified high-flow oxygen delivered by the HFNC helps secretion clearance, decreases airway inflammation, and decreases energy requirement, especially in acute respiratory failure. Delivering high-flow oxygen improves the work of breathing and matches the high peak inspiratory flow demands in patients with respiratory illness.[1011]

In this trial, we included males and females with a wide age range between 10 and 40 years to make the results significant, and we then compared HFNC with the standard apnoeic oxygenation during FB removal using a rigid bronchoscope. The results showed that oxygen saturation was higher in the HFNC group that helped reduce the number of bronchoscopy attempts and the total procedure duration. The end-tidal carbon dioxide levels recorded at the time of mask ventilation at the end of the procedure were significantly low in the HFNC group compared to the apnoeic group.

Our results came in accordance with a previous study, which suggested that HFNC oxygen therapy might be better in lowering the occurrence of severe hypoxaemia.[12] Lucangelo et al.[1] compared the delivery of 50% oxygen before and during bronchoscopy using either a HFNC (40 or 60 L/min) or a Venturi mask in patients undergoing bronchoscopy. Patients who received 60 L/min via the HFNC maintained higher PaO2 values, higher arterial–alveolar oxygen tensions, and higher PaO2/FiO2 ratios both during and after the procedure.

An earlier randomised trial conducted on patients with hypoxaemia undergoing flexible fiberoptic bronchoscopy for diagnosis of pulmonary lesions reported HFNC as a well-tolerated, effective technique for oxygen delivery in those patients. Oxygen was delivered during the procedure through a nasal cannula with an inspiratory flow rate of 40 L/min, and the FiO2 was kept at 0.6 throughout and 30 min after the procedure.[4]

Min and colleagues applied preoxygenation with HFNC for 3 min before and during induction of general anaesthesia for rigid bronchoscopy. The flow of oxygen was set at 50 L/min, and the FiO2 was kept at 100%. The investigators reported HFNC as a better tool for oxygenating patients during rigid bronchoscopy compared to the conventional method of preoxygenation.[13]

Vourc’h et al.[14] enroled adult patients with respiratory failure requiring intubation and divided them to receive either HFNC oxygen (100% FiO2, 60 L/min) or 15 L/min oxygen delivered by a face mask. Contrary to our results, the authors concluded that using the HFNC without discontinuation during an apnoeic period was not effective in preventing desaturation regardless of the severity of respiratory distress.

A more recent study compared HFNC to standard oxygen therapy in immunocompromised patients with acute respiratory failure. The results showed no difference between the two groups regarding the 28-day mortality. Despite improvement in PaO2/FiO2 ratio, HFNC failed to show a better outcome as regards the percentage of intubation, length of ICU stay, rate of infection, patient comfort, or dyspnoea scores.[15]

The current trial found that PO atelectasis assessed by chest radiogram was less frequent in the HFNC group, which came in accordance with previous studies assessing PO hypoxia and basal atelectasis in patients who had HFNC oxygenation with some favourable results. Stephan and colleagues found that the use of HFNC in patients with hypoxaemia following cardiac surgery was effective in decreasing the incidence of basal atelectasis and the need for reintubation.[16]

Serial arterial blood gases analysis would have shown a great value in comparing PaO2 and partial pressure of carbon dioxide (PaCO2) between the two groups, but we did not use this tool as it is not used in the protocol of our institute.

CONCLUSION

In conclusion, HFNC was superior to apnoeic oxygenation in achieving oxygen saturation above 92% in patients undergoing rigid bronchoscopy for FB removal. Compared to the standard apnoeic oxygenation, the use of HFNC provided higher oxygen saturation and lower end-tidal carbon dioxide levels. Also, it provided a lesser interruption of the procedure and fewer attempts at bronchoscopy trials.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.