High Flow Oxygen During Operative Hysteroscopy.

High-flow nasal cannula (HFNC) oxygen therapy represents an open circuit ventilation system that uses flows up to 70 L/min of 100% oxygen through Optiflow THRIVETM (transnasal humidified rapid-insufflation ventilatory exchange) device (Fisher and Paykel Healthcare Ltd, Auckland, New Zealand).

Compared to conventional low-flows oxygen therapy systems, the heating and humidification of the flow facilitate tolerability by the patient, ensure an efficient muco-ciliary function and allow to reach higher and more stable inspiratory fractions of oxygen.

The system produces a flow-dependent effect of continuous positive airway pressure of about 1 cmH20 for every 10L/min increase in oxygen flow and a reduction of the dead space, having the potential to increase the alveolar volume thus improving gas exchanges.

The resulting reduction in respiratory rate and the improvement of thoracoabdominal synchrony minimize the onset of respiratory fatigue in the patient.

Initially introduced for the treatment of acute hypoxemic respiratory failure, HFNC oxygen therapy is increasingly used also in anesthesia as the only airways management technique during short-term procedures performed under sedation (e.g. digestive endoscopy) or under general anesthesia (such as microlaryngoscopies).

Compared to low-flows oxygen systems, the greater CO2 elimination capacity could reduce the incidence of hypoventilation and apnea episodes potentially associated with patient sedation.

Operative hysteroscopy is a brief procedure, normally lasting <30 minutes, usually performed in a day-hospital regime under procedural analgo-sedation, with conventional ventilatory support through a face mask/laryngeal mask.

The primary objective of this study is the comparison of the rate of success in airways management via the THRIVE system versus positive pressure ventilation by I-gel laryngeal mask during 180 hysteroscopies performed under general anesthesia.

Secondary objectives are: the comparison of intraoperative and postoperative complications between the two groups.

Materials and methods Patient monitoring throughout the procedure will include automatic noninvasive, intermittent (every 5 minutes) blood pressure monitoring, 3-leads electrocardiogram, peripheral oxygen saturation (SpO2), bispettral index (BIS), transcutaneous capnography with TCM5 Radiometer monitor.

Upon arrival in the operating room, the patient will assume the lithotomic position; a peripheral venous cannula will be placed at the level of the hand or forearm and it will begin the infusion of Ringer Lactate solution 3 ml/kg/h. Omeprazole 40 mg and dexamethasone 4 mg will be administered before the procedure as standard internal practice.

In the THRIVE group, dedicated Optiflow THRIVETM nasal cannulas will be positioned for 100% high flow oxygen therapy at an initial flow of 30 L/min. After induction of anesthesia and throughout the procedure the flow of oxygen will be increased to 70 L/min.

In the control group (CONTROL) after induction of anesthesia the patient will be mechanically ventilated through an I-Gel laryngeal mask.

General anesthesia, as per normal clinical practice, will be induced and maintained by target-controlled infusion (TCI) of propofol (7 mcg/Kg) by Orchestra Infusion system (Fresenius Kabi) and fentanyl 1.5 mcg/kg and in both groups. Propofol infusion rate will be between 3-4 mcg/ml for propofol, so to maintain a level of sedation monitored via BIS between 40-50. Paracetamol 1g, ondansetron 4 mg and ketorolac 30 mg will also be administered as per normal clinical practice.

In case of desaturation episodes, defined as a SpO2< 94% or increases in tcCO2 above 65 mmHg, the patient will be assisted with positive pressure ventilation through a facial or laryngeal mask or oro-tracheal intubation based on clinical judgment of the anesthesiologist.

At the end of the hysteroscopy the patient will be transferred to the Post-Anesthesia Care Unit (PACU) for 3 hours of post-procedural monitoring as required by internal protocol for day-surgery procedures. Standard monitoring will include non-invasive, intermittent (every 15 minutes) blood pressure monitoring, 3-leads electrocardiogram and peripheral oxygen saturation (SpO2).

One hour after the end of the procedure and before discharge from the PACU the patient will be asked to quantify the perceived degree of dyspnea with the Borg dyspnoea score and the degree of comfort using a Visual Numeric Scale (VAS) ranging from 0 to 10. In case of SpO2<94%, additional oxygen will be administered by means of a Venturi-type face mask with FiO2 40%. Discharge from the PACU will require an Aldrete score of 9-10.

Statistics The data will be analyzed according to an intention-to-treat principle. Clinical and demographic characteristics of the sample will be described through descriptive statistical techniques. Continuous quantitative variables with normal distribution will be reported as mean and standard deviation; as median and interquartile range the non-normal variables.

Categorical variables and missing data will be presented as absolute values and percentage, n (%).

Continuous variables will be compared with the Student t-test in case of normal distribution or, if not, with the Mann-Whitney test for independent samples. The normality of the distribution of the variables will be verified graphically by histograms and with the Shapiro-Wilk test.

Continuous variables with repeated measurements will be compared with a mixed-effect linear regression model. The type of airway management used and the timing at which the measurements will be performed will be considered as fixed effects; a random effect related to the patient will also be added.

The assumption of normality of the residues of the model will be graphically verified by histograms and Q-Q plots. The P-values will be obtained with the likelihood ratio test of the complete model with the effect in question with respect to the model without the effect. The estimates of the predictors used in the model will be reported with the relative 95% confidence intervals and P-values obtained with the Wald test.

In case of failure of the technique, defined as a transcutaneous concentration of carbon dioxide (tcCO2) > 65 mmHg and/or SpO2 < 94% in any of the study groups, a time-to-event (Kaplan-Meier) analysis will be carried out that takes into account the different duration of the procedures and therefore the different risk ranges for the possible onset of hypercapnia and / hypoxemia.

Differences between categorical data will be reported in terms of relative risk or risk differences, 95% confidence intervals and p-values based on the Χ2 test or fisher's exact test.

The significance level will be set for α < 0.05. All analyses will be performed with the statistical software R version 4.1.2 (R Foundation for Statistical Computing, Austria).

Continuous variables with repeated measurements will be compared with a mixed-effect linear regression model. Normality of distribution will be verified with the Shapiro-Wilk test. Continuous variables will be compared with Student t - or Mann Whitney test; categorical variables with the Chi-square test.

Sample size calculation In 2021, Kim et al. compared the safety of high flow oxygen therapy to conventional low-flow oxygen support for gastrointestinal endoscopic procedures performed under sedation reporting minimum values of peripheral saturation detected by pulse-oximeter higher in the high-flow group (99.8%) compared to low-flow oxygen therapy (95.1%).

To our knowledge, no study has compared these two methods during operative hysteroscopies under procedural sedation.

Based on literature and on the results from an internal registered pilot study in which we investigated the success rate of THRIVE as unique airway management technique in this setting, assuming a similar success rate of 95% for THRIVE and laryngeal mask ventilation, for a unilateral confidence interval of 95% and a 90% test power, we estimated a minimum sample of 82 patients per group to test a non-inferiority limit of 10%. The sample was increased to 90 patients for group to take into account any dropouts.