Pre‐oxygenation using high‐flow nasal oxygen in parturients undergoing caesarean section in general anaesthesia: A prospective, multi‐centre, pilot study

Parturients undergoing caesarean section in general anaesthesia have an increased risk of desaturating during anaesthesia induction. Pre‐ and peri‐oxygenation with high‐flow nasal oxygen prolong the safe apnoea time but data on parturients undergoing caesarean section under general anaesthesia are limited. This pilot study aimed to investigate the clinical effects and frequency of desaturation in parturients undergoing caesarean section in general anaesthesia pre‐ and peri‐oxygenated with high‐flow nasal oxygen and compare this to traditional pre‐oxygenation using a facemask.

high-flow nasal oxygen group vs 80% (15) in the facemask group (p = .308). No signs of regurgitation, in any parturient, were seen.
Conclusion: Pre-and peri-oxygenation with high-flow nasal oxygen maintain adequate oxygen saturation levels during induction of anaesthesia also in parturients.
Regurgitation of gastric content did not occur in any parturient and no other safety concerns were observed in this pilot study.

Editorial Comment
High-flow nasal oxygen supplementation in pregnant women undergoing caesarean section with general anaesthesia is an alternative method to this purpose. This pilot study showed feasibility for maintaining adequate oxygenation without adverse effects. Randomised clinical trials are warranted to confirm these results.

| INTRODUCTION
Caesarean section is one of the most frequently performed surgical interventions globally with over a million procedures done yearly in the United States only. 1 The majority of all caesarean sections are performed in neuraxial anaesthesia. General anaesthesia is typically avoided because of potential risks, the most feared being difficult tracheal intubation with subsequent hypoxia and, in worst case, death. 2 Maternal hypoxia also negatively affects the foetus with an increased risk of morbidity and need for neonatal intensive care. 2,3 Pregnant women desaturate quickly during apnoea. Physiological changes during pregnancy limit the capacity to store oxygen due to a smaller functional residual capacity. 4 Furthermore, pregnant women have an increased oxygen consumption, 5 thus depleting oxygen stores quickly. Due to mucosal swelling during pregnancy and labour with subsequent increase in Mallampati score, these patients more often present with a difficult airway 6 and failed intubation is not uncommon. 7 Previous studies investigating pregnant women undergoing general anaesthesia have shown that the incidence of desaturation during induction of anaesthesia is high. 8,9 A prospective observational study from 2020 investigating pregnant women undergoing general anaesthesia for non-elective caesarean sections showed that 19% desaturated during induction of anaesthesia whilst severe hypoxemia, defined as SpO 2 ≤90%, was observed in 9.4% of the cohort. 8 Pre-oxygenation prolongs the time until desaturation during apnoea. 10 Lately, high-flow nasal oxygen (HFNO) has been investigated in trials and used clinically to prolong the safe apnoea time. 11,12 Previous studies have shown that HFNO generates an oxygen reserve during pre-oxygenation comparable to the standard tight fitting face mask. 13,14 Pre-oxygenation with HFNO makes a seamless transition from pre-oxygenation to apnoeic oxygenation possible. There are studies showing that apnoeic patients can stay well oxygenated for up to half an hour with HFNO as only source of oxygenation after preoxygenation. 11,12 Studies investigating end-tidal oxygen levels during simulation of pre-oxygenation in volunteer parturients, not undergoing anaesthesia or surgery, have shown worse results from preoxygenation with HFNO compared to traditional facemask. [15][16][17] Other studies have shown promising results, clinically, when HFNO has been used both as pre-and peri-oxygenation until a definite airway has been established. 18,19 We hypothesised that pre-and peri-oxygenation using HFNO would be a feasible method to use, maintaining oxygen saturation levels at least equivalent to pre-oxygenation using a traditional facemask, also in parturients. The aim of this study was to investigate the clinical effects and frequency of desaturation in parturients undergoing caesarean section in general anaesthesia pre-and peri-oxygenated with HFNO and compare this to traditional pre-oxygenation using a facemask, in a broader context, including multiple participating centres and limited criteria for exclusion.

| Ethics
Ethical approval for this prospective, non-randomised, multi-centre pilot study (Dnr 2019-05399) was provided by the Swedish ethical board, Uppsala (Chairperson Per-Erik Nistér) on 21 January 2020. The study was registered at clinicaltrials.gov (NCT04711317). Written and oral informed consent was obtained from all included patients. Exclusion criteria were category 1 caesarean section (immediate threat to the life of parturient or foetus), BMI > 45 kg m À2 (pregnant weight), need for non-invasive ventilation before anaesthesia and inability to give consent.

| Experimental design
All eligible parturients were asked to participate in the intervention group, hence being pre-oxygenated with HFNO (Optiflow™, Fisher & Paykel Healthcare, Auckland, New Zealand). Parturients who declined to participate were asked for permission to collect background data as well as physiological data during induction of anaesthesia when pre-oxygenation was performed with traditional, tight fitting, facemask. This group of women constituted the control group.
Background data and comorbidity were collected either from the parturient before surgery or from charts.
This study was originally designed to only include parturients preand peri-oxygenated with HFNO to investigate the feasibility of this method also in this population. In Sweden, the predominant amount of all caesarean sections is performed under neuraxial anaesthesia why the use of general anaesthesia is rare. In order to recruit as many parturients as possible we decided to also include a control group to the study, consisting of parturients declining to participate in the intervention group. Early in the recruitment process we changed the primary outcome from being a comparison in lowest median SpO 2 between the groups to instead comparing the proportion of parturients desaturating below 93% from start of pre-oxygenation until 1 min after intubation. The change was made because we thought that this was a more clinically relevant outcome. It also enabled us to compare our data to previous studies investigating the frequency of oxygen desaturation amongst parturients undergoing caesarean section in general anaesthesia.

| Experimental protocol
On arrival to the operating theatre all parturients were monitored with ECG, pulse oximetry and non-invasive blood pressure measurement (Philips Intellivue MX800 with X2 or X3, Amsterdam, The Netherlands). They were supplied with at least one intravenous line where an infusion of a standard electrolyte solution was started. The parturient was then placed supine, in a slightly reverse Trendelenburg position, on the operating table. A wedge pillow was put under the right hip allowing for a pelvic tilt to the left. General anaesthesia was performed with a rapid sequence induction technique. In the HFNO group pre-oxygenation consisted of 100% oxygen administered for at least 3 min at a flow of 30-50 L min À1 , as tolerated. The parturients were not specifically instructed to breathe with a closed mouth but to keep it open or closed at their own convenience. Once apnoea occurred the flow was increased to 70 L min À1 and an open airway was maintained using chin lift or jaw thrust until laryngoscopy was initiated and tracheal intubation was completed. The ventilator was set to deliver 100% oxygen during pre-oxygenation in order to flush the anaesthetic tubes prior to intubation. In the facemask group parturients were pre-oxygenated according to standard routine at each centre. In most cases this consisted of delivering pure oxygen at a flow of 10 L min À1 via a tight fitting facemask connected to an anaesthetic circuit. All participating centres were instructed to pre-oxygenate all parturients for a minimum of 3 min.
Drugs and doses administered during induction of anaesthesia were determined by the anaesthetist in charge.
During the procedure apnoea time, intubation time and lowest SpO 2 from start of pre-oxygenation until 1 min after intubation were recorded as well as the lowest mean arterial pressure and lowest heart rate during the induction of anaesthesia. Apnoea time was defined as start of apnoea, assessed clinically, until a carbon dioxide trace was visible on the capnograph. Intubation time was defined as start of laryngoscopy, when the laryngoscope passed the teeth, until a carbon dioxide trace was visible on the capnograph. Collection of parturient data ended 1 min after intubation. Apgar score from the newborn was noted at 1, 5 and 10 min following delivery.

| Primary endpoint
Primary outcome of this study investigated differences in the proportion of parturients desaturating below 93% from start of preoxygenation until 1 min after intubation between the group pre-and peri-oxygenated with HFNO and the group pre-oxygenated with facemask.

| Secondary endpoints
As secondary outcomes, we investigated differences between the two groups in end-tidal gas concentrations in the first breath after intubation as well as the proportion of parturients with signs of regurgitation.

| Statistical analysis
Since this study was conducted as a pilot study to show a proof of concept no power analysis was done before study start. When  initiating this study there were no available data on HFNO as pre-and peri-oxygenation in parturients undergoing general anaesthesia and this study was designed to assess the feasibility of the concept. We, therefore, decided to conduct a small investigating pilot trial before potentially conducting a larger, randomised, study.
We estimated that 25-30 parturients pre-oxygenated with HFNO would be sufficient to show whether this is a feasible method to use, without major safety concerns, also to this particular patient population.
Continuous data were analysed using an unpaired two-sided t-test or a Mann-Whitney U-test depending on the distribution.
Categorical data were assessed using a Chi-square test or, in case of sample size assumption violation, a Fisher's exact test. A p value <.05 was considered statistically significant. All statistical analyses were performed using SPSS Statistics ® 28 (IBM ® , Armonk, NY, USA).

| Study population
Parturients were enrolled between May 2021 and October 2022. A total of 34 women were included in the study, 25 were preoxygenated with HFNO and 9 with standard tight fitting facemask. At one centre, one parturient in the control group received a low flow of oxygen (10 L min À1 ) via a standard nasal cannula below the facemask allowing for continuous oxygenation during the apnoeic period. This low-flow apnoeic oxygenation added to one patient was performed since this was clinically routine at this centre. Thirty-two parturients underwent elective caesarean section. Two parturients in the HFNO group had category 3 caesarean section performed, both needed to deliver early but there was no immediate risk to the mother or foetus.
One of these women was in labour but converted to caesarean section due to failure to progress. No included parturient was lost to follow-up or excluded due to protocol violence. Eight parturients who were assessed for eligibility were excluded (Figure 1).

| Patient and airway characteristics
Parturients in the HFNO group had a higher BMI than the control group, 31.4 ± 4.7 kg m À2 versus 27.7 ± 3.1 kg m À2 ; p = .034. No other differences in patient or airway characteristics were seen between the groups (Tables 1 and 2). The most common indication for general anaesthesia was parturient request followed by coagulopathy (Table 1).

| Conditions at intubation
Conditions at intubation are shown in Table 3. Duration of pre-oxy- Apgar score of the newborn was followed until 10 min after delivery. No difference in the score depending on the method of preoxygenation was demonstrated (see Table S1). The drugs of choice did not differ between the groups (see Table S1). In most cases, the regime consisted of an induction drug and a neuromuscular blocking drug whilst some parturients also received an opioid.
T A B L E 2 Airway characteristics of 34 parturients undergoing caesarean section in general anaesthesia pre-and peri-oxygenated with high flow nasal oxygen or facemask.

| Primary endpoint
wNo parturient in either group had a peripheral oxygen saturation below 93% from start of pre-oxygenation until 1 min after intubation (Figure 2).
The lowest SpO 2 measured during induction of anaesthesia in the whole cohort was 97%. The lowest median SpO 2 in both groups was 100%.

| Secondary endpoints
There was no difference in mean end-tidal oxygen concentrations in the first breath after tracheal intubation between the two groups (Table 3). However, more parturients pre-oxygenated with facemask had end-tidal oxygen levels below 70%, 3 (33.3%) in the facemask T A B L E 3 Secondary outcomes and conditions at intubation in 34 parturients undergoing caesarean section in general anaesthesia pre-and peri-oxygenated with high flow nasal oxygen or facemask.
High flow nasal oxygen (n = 25) Facemask (n = 9) p value F I G U R E 2 Oxygenation until 1 min after intubation. Lowest SpO 2 from start of preoxygenation until 1 min after intubation and apnoea time is plotted for every parturient preoxygenated with high-flow nasal oxygen (n = 25) or facemask (n = 9). High-flow nasal oxygen (blue squares) and facemask (red dots). Desaturation was defined as SpO 2 < 93% (dotted line).
group vs 1 (4.0%) in the HFNO group (p = .048) (Figure 3). No differences between the groups in end-tidal carbon dioxide levels in the first breath after intubation could be seen and no parturient in this study showed signs of regurgitation (Table 3).

| DISCUSSION
In this prospective, non-randomised, multi-centre study we demonstrate that pre-oxygenation with HFNO in parturients undergoing caesarean section in general anaesthesia is efficient in maintaining oxygen saturation, at least comparable to pre-oxygenation with a traditional tight-fitting facemask. Also, there were no signs of an increased risk of regurgitation.
No woman desaturated below 97% during induction of anaesthesia. A previous study investigating the same population preoxygenated with traditional facemask has shown an incidence of desaturation of 16% 9 with an even higher incidence during nonelective caesarean sections. 8  This is, to our knowledge, the first European study investigating pre-oxygenation using HFNO in parturients undergoing caesarean section in general anaesthesia. In contrast to previous studies investigating HFNO as pre-and peri-oxygenation in parturients undergoing general anaesthesia, this study was conducted at multiple centres, mainly to increase the generalisability of our results but also to enhance patient recruitment. Also, most parturients were eligible for F I G U R E 3 End-tidal oxygen concentrations (EtO 2 ) in first breath after intubation. Lowest endtidal oxygen concentration in the first breath after intubation and apnoea time plotted for every parturient pre-oxygenated with high-flow nasal oxygen (n = 25) and facemask (n = 9). High-flow nasal oxygen (blue squares); facemask (red dots). End-tidal oxygen concentrations <70% (dotted line) were seen in one parturient (4.0%) preoxygenated with high-flow nasal oxygen and three parturients (33.3%) pre-oxygenated with facemask (p = .048).

| LIMITATIONS
There are some limitations to our study. The study was not randomised. Caesarean section in general anaesthesia is a low-frequency event in Sweden. In order to include a sufficient number of parturients to evaluate the concept of HFNO as pre-and peri-oxygenation we initially decided to include all patients in the intervention group.
During patient enrolment, we decided to also include a control group consisting of parturients not wanting to participate in the intervention group. This was done to increase the sample size but also to enable a comparison of the results from the HFNO-group. There could therefore be confounding factors in the control group of our study. The study could also suffer from a selection bias since women in the control group constituted of those not wanting to participate in the HFNO group which could introduce skewness in data. Other than a minimum time of 3 min, the duration of pre-oxygenation was not set in the study protocol, this to mimic the clinical setting as closely as possible. Some patients, in both groups, therefore received a significantly longer pre-oxygenation. Even though previous studies have shown that an additional effect from a pre-oxygenation extending beyond 3 min is limited this could be considered a weakness.
This study excluded women in need of category 1 caesarean sections. These patients are probably more prone to desaturate during induction of anaesthesia and could therefore benefit from a procedure that prolongs the safe apnoea time. Lastly, this was a relatively small, pilot trial, with no women presenting with a difficult airway prolonging the apnoea time, circumstances where the benefits from using HFNO probably are the greatest. Further studies would therefore be of interest.

| CONCLUSION
In this study we show that HFNO as pre-and peri-oxygenation in pregnant women undergoing caesarean section in general anaesthesia could be an adequate technique to use and comparable to facemask.
In our study, we did not see an increased risk of regurgitation amongst parturients pre-oxygenated with HFNO, although these results are limited by the size of the population.