Study design and setting
We will conduct this prospective multicentre randomised trial as per the recommendations for interventional trials (SPIRIT, Fig. 1 and Additional file 1) . The final reporting of this trial will be in accordance with the Consolidated Standards of Reporting Trials (CONSORT) statement.
The OTHER trial will be conducted across three Australian hospitals: The Queen Elizabeth Hospital, and Royal Adelaide Hospital, South Australia and John Hunter Hospital, New South Wales. A total of 132 patients will be recruited. ERCP procedures are unique in terms of their presentation. A good proportion of them are undertaken as semi-urgent cases (unlike other pure electively planned cases) as inpatients and they may not get the opportunity to present to a pre-admission clinic. We are planning to recruit these patients at least two hours before the procedure or whenever they are being assessed for anaesthesia. During this time, the patient information sheet will be provided. The recruitment and consenting is done by one of the investigators or by the anaesthetists assessing these patients. The pre-screening process would be based on the inclusion criteria mentioned below.
Participants will be randomly assigned to either the low flow nasal oxygen of the HFNC group at a ratio of 1:1 (Fig. 2). The randomisation scheme will be generated by the Clinical trials division of the pharmacy department at The Queen Elizabeth Hospital. To ensure equal distribution of the intervention arm, stratification is done in specific blocks to predetermined numbers known only to the clinical trials division. This will be revealed only at the end of the trial. The random numbers and the group assignment will be supplied in sealed envelopes and handled only by the principal investigator. A set of 25 envelopes will be dispatched and used at each site, and additional set of envelopes will be dispatched in blocks of 20 once the initial 25 are used. The envelopes will be opened just before commencing ERCP and groups are allocated to the treatment intervention. Participants and the investigators are not blinded to the allocation. The patient information sheet contains descriptions of the trial. Diagrams of high flow as well as low flow nasal cannula are depicted in the sheet. Further, when we commence high flow nasal cannula we are obliged to explain the patient that they will feel a high flow of oxygen coming through their nostrils. These are the reasons why we couldn’t blind the participants to the intervention.
Data will be collected only by the investigators. The data analysts will be blinded to the intervention.
Adults (aged over 18 years) fulfilling any of these criteria: ASA 3, or 4, obesity (BMI above 30 kg/m2), obstructive sleep apnoea diagnosed either by polysomnography, being treated with CPAP for OSA, suspected OSA based on STOP BANG score >/3*
*STOP BANG is an acronym which abbreviates Snoring, Tiredness, Observed choking/gasping, blood Pressure elevation, BMI increase, Age, Neck size, male Gender.
Participants fulfilling any of the below criteria will be excluded: 1)Deemed “difficult airway” and/or difficult intubation based on clinical judgement and known previous difficult airway; 2) Severe cardio-respiratory compromise or any other indications that necessitate the procedure to be done under general anaesthesia with endotracheal tube; 3) Patients judged to be at significant risk of pulmonary aspiration. Risk assessment will be based on patient history (focussing particularly on risk factors for aspiration) and physical examination. Possible risk factors for aspiration include: Increased gastric content, delayed gastric emptying, including lap band in situ, lack of fasting (less than six hours’ solids and two hours for clear fluid), Increased regurgitation risk: Uncontrolled or symptomatic gastro-oesophageal reflux, oesophageal strictures, Zenker Diverticulum and achalasia, Laryngeal incompetence due to cerebral infarct, head injuries, neuromuscular disorders (Parkinson’s disease, Gullian Barre), muscular dystrophies (cerebral palsy, cranial neuropathies); 4) Emergency surgery and any other criteria warranting general anaesthesia with an endotracheal tube
Intervention and comparison
Patients will be randomly allocated (computer generated randomisation) to either the low flow nasal oxygen (L) group (group L; n=66) or the high-flow nasal oxygen (H) group (group H; n=66). In group L, the procedure will be performed under deep sedation and analgesia, with supplemental oxygen 4 L/min via regular nasal cannula and a further supplemental oxygen source with a flow rate of 4 L/min administered through an extension tubing attached to the mouth guard. In group H, the procedure will be performed under deep sedation and analgesia similar to Group L, with oxygen delivered through a HFNC. This will be accomplished using the Optiflow THRIVE (Transnasal Humidified Rapid Insufflation Ventilatory Exchange) device (Opti-Flow, Auckland, New Zealand). Flow rate through the cannula will be commenced at 30 L/min and fractional inspired oxygen concentration will be set at 100%. The flow will be gradually increased after the administration of sedative agents, and will be maintained at 50 L/min during the procedure. The flow rate may be either increased up to 70 L/min if necessary or decreased to 30 L/min if the patient does not tolerate the higher flow rate. (50 L/min is very well tolerated in even very lightly sedated patients.)
Standard monitoring will be applied including continuous electrocardiogram, blood pressure automatically measured every 3 minutes, transcutaneous capnography, and pulse oximetry. The procedures will be done in either lateral or prone positions. A standard sedation technique will be applied for both groups. Sedation is to be provided by titrated doses of fentanyl 0.5-1.0 mcg/kg as required, and a propofol target controlled infusion (using the Marsh model) commencing at an initial plasma target of 1.5 to 2.0 mcg/ml and titrated up or down during the procedure, between target values of 1 to 4 mcg/ml based on satisfactory depth of anaesthesia and frequency of ventilation (8 – 14 breaths per min). Administration of further fentanyl top up doses (25 mcg) will be at the anaesthetists discretion. In all cases airway obstruction needs to be avoided.
If it is deemed necessary that during the intervention phase, if the participants’ clinical condition warrants an advanced airway such as tracheal intubation, it will be undertaken and they will be retained in an intention-to treat analysis.
The occurrence of hypoxia, defined by any event of SpO2 (oxygen saturation measured by pulse oximetry) <90% of any duration: Proportion of patients experiencing desaturation below 90% at any time during the procedure will be compared between two groups
- Number of events of hypoxia, defined as desaturation < 90%. The mean number of events during the procedure will be compared between the two groups.
- Lowest recorded SpO2 during the procedure
- Transcutaneous CO2(carbon-di-oxide). Maximum value recorded and average value during the case. The mean values during the procedure will be compared between the two groups.
- Lowest recorded SpO2 % during the procedure. The mean values during the procedure will be compared between the two groups.
- Requirement of minor airway manoeuvres: jaw lift/jaw thrust, nasopharyngeal airway insertion. Proportion of patients requiring these manoeuvres will be compared between two groups
- Requirement of major airway manoeuvres: bag mask ventilation, endotracheal intubation. Proportion of patients requiring these manoeuvres will be compared between two groups
- Arrhythmia (any change in cardiac rhythm observed on the electrocardiogram). Proportion of patients manifesting arrhythmia during the procedure will be compared between two groups
- Total fentanyl dose. The mean values, (the doses in micrograms) during the procedure will be compared between the two groups.
- Requirement of antispasmodic agent. Proportion of patients requiring this medication will be compared between two groups
- Total duration of procedure (starting from sedation until leaving the suite). The mean duration in minutes will be compared between the two groups.
- Duration under sedation/anaesthesia: Measured from the time of induction till eye opening. The mean duration in minutes will be compared between the two groups.
- Successful completion of the procedure: Yes/No. Proportion of patients fulfilling this criteria will be compared between two groups
- Early complications - elicited as patients leave Recovery
- Dry mouth/nose/throat: Binary outcome (constant pain or discomfort in the mouth/nose/throat). Proportion of patients experiencing this adverse event will be compared between two groups
- Sensation of abdominal bloating: Y/N. Proportion of patients experiencing this adverse event will be compared between two groups
- Patients’ satisfaction score on leaving recovery: 5 points numerical rating scale:
- Very satisfied (5), somewhat satisfied (4), Neither satisfied nor dissatisfied (3), Somewhat dissatisfied (2), Very dissatisfied (1). Proportion of patients at particular threshold will be compared between two groups
- Cost analysis of the consumables: low flow vs High flow. This will be analysed for only the consumables required during the procedure for an average case, and compared across two groups. The information will be obtained from our clinical pharmacy and anaesthetic nursing department.
Sample size calculation
A 21.4% percent incidence of hypoxia (SPO2<90% for 15 seconds) was noted in an Australian study during ERCP with propofol sedation in older patients (20). Assuming a 16% reduction in hypoxic events by employing the HFNC technique, (given the rate of hypoxia as 21.4% in the sedation group and an estimated 5.4% in the HFNC group), with a power of 80% and an alpha error of 0.05, a total of 132 patients would be required (66 in each arm).
Mean and standard deviation values will be estimated for continuous outcomes while frequency and percentage will be computed for binary outcomes. 95% Confidence intervals around the point estimate will be calculated where appropriate for the primary and secondary outcomes. Descriptive statistics will be used to present the results. P-value <0.05 to be considered significant. Analyses will be intention-to-treat from randomisation. All randomised cases will be included in the analyses, regardless of missing data. As the data capture is only limited to few hours after the intervention and the investigators are directly involved in the conduct of the study, we anticipate very less missing data. A subgroup analyses will be attempted (if feasible) for the high BMI and OSA groups combined.
Strengths and limitations of this study
Our study is the first multi-centre randomised controlled trial comparing low flow versus high flow nasal oxygen therapy for improving oxygenation in high risk patients for Endoscopic retrograde cholangiopancreatography. There are standardised and objective endpoints. Also patient reported outcome measures are explored. The application of transcutaneous carbon dioxide measurements helps to overcome the limitation of expired carbon dioxide monitoring during ERCP. The nature of the study precludes blinding of the participants and the anaesthetists, which may contribute to bias and influence the results
Quality control and data monitoring
The investigators and anaesthetic nursing staff will undergo special training on using and troubleshooting the transcutaneous CO2 equipment. Only data on respiratory, haemodynamic and perioperative outcome endpoints will be collected during the procedure and it is not time bound. Only one set of data will be collected in post anaesthesia care unit when they leave the area. The data will be de-identified when entered in to excel sheet for analysis. No blood or tissue samples will be collected. The original data collection will be kept for cross checking. The data forms will be securely placed in our department’s locked filing cabinets. Any serious adverse events across any sites will be notified within 24 hours to the principal investigator. Patient recruitment and data quality will be regularly checked by the investigators across the trial centres. The study involves comparing established standard practices and hence deemed as low risk to be specifically monitored by a data monitoring committee.
Ethics and dissemination
This study has been approved by the Central Adelaide Local Health Network Human Research Ethics Committee (Version 4, dated 31/10/2018; Approval date 25/09/2018. Ethics approval number: Q20180807) as well as by the Hunter New England Local Health District Research governance (NSW REGIS No: 2019PID00554, SSA reference No: 2019/STE0047). This study was registered a month after the first patient was recruited at www.ANZCTR.org.au CTRN12619000397112 on March 12, 2019. and registered at Australian New Zealand Clinical Trials Registry (ANZCTR). ANZCTR Registration number: ACTRN12619000397112. The results of the study will be disseminated through peer-reviewed publications and national/international conference presentations. Protocol amendments will be immediately notified to all centres.