Automated oxygen administration versus conventional oxygen therapy after major abdominal or thoracic surgery: study protocol for an international multicentre randomised controlled study

Abstract

Introduction

Hypoxemia and hyperoxia may occur after surgery with potential related complications. The FreeO₂ PostOp trial is a prospective, multicentre, randomised controlled trial that evaluates the clinical impact of automated O₂ administration versus conventional O₂ therapy after major abdominal or thoracic surgeries. The study is powered to demonstrate benefits of automated oxygen titration and weaning in term of oxygenation, which is an important surrogate for complications after such interventions.

Methods and analysis

After extubation, patients are randomly assigned to the Standard (manual O₂ administration) or FreeO₂ group (automated closed-loop O₂ administration). Stratification is performed for the study centre and a medical history of chronic obstructive pulmonary disease (COPD). Primary outcome is the percentage of time spent in the target zone of oxygen saturation, during a 3-day time frame. In both groups, patients will benefit from continuous oximetry recordings. The target zone of oxygen saturation is SpO₂=88%–92% for patients with COPD and 92%–96% for patients without COPD. Secondary outcomes are the nursing workload assessed by the number of manual O₂ flow adjustments, the time spent with severe desaturation (SpO₂ <85%) and hyperoxia area (SpO₂ >98%), the time spent in a hyperoxia area (SpO₂ >98%), the VO₂, the duration of oxygen administration during hospitalisation, the frequency of use of mechanical ventilation (invasive or non-invasive), the duration of the postrecovery room stay, the hospitalisation length of stay and the survival rate.

Ethics and dissemination

The FreeO₂ PostOp study is conducted in accordance with the declaration of Helsinki and was registered on 11 September 2015 (http://www.clinicaltrials.gov). First patient inclusion was performed on 14 January 2016. The results of the study will be presented at academic conferences and submitted to peer-reviewed journals.

Trial registration number

NCT02546830.

Strengths and limitations.

• The FreeO₂ PostOp trial is the first and largest clinical evaluation of automated oxygen titration over standard of care in patients undergoing major surgeries.

• As automated oxygen titration is not the standard of care for postoperative patients, the study was designed as a superiority study.

• The use of continuous and non-averaged SpO₂ values monitoring in all groups of patients enables precise oxygenation evaluation during the entire study period.

• One limitation of the study is that investigators are aware of the inclusion group, while blinding is difficult in studies with respiratory support.

• Second limitation could be that the oxygenation status assessment would have been more precise using arterial blood gases.

Introduction

Background and rationale

Current standards for prescribing oxygen recommend providing adequate flows to correct hypoxemia and avoid hyperoxia.^{1 2} While deleterious effects of hypoxemia are well known, the potential harmful effects of hyperoxia are underappreciated. Yet, hyperoxia may increase mortality in patients with severe chronic obstructive pulmonary disease (COPD)^3–5 and may cause cardiac and neurological adverse toxicities in certain situations.^6–8 Precise control of O₂ flows is difficult to achieve in clinical practice and is time-consuming.⁹

The FreeO₂ system (Oxynov, Quebec, Canada) is an innovative device, developed in collaboration between our researchers from Brest-France, Quebec-Canada University Hospitals and Oxynov Inc., an R&D spin-off from Laval University-Quebec. FreeO₂ is a closed-loop device that automates oxygen administration to spontaneously breathing patients in response to pulse oximetry (SpO₂) continuous measurements.¹⁰ Automated O₂ administration allows to maintain constant SpO₂ within a predetermined range using variable O₂ flows, as opposed to manual O₂ administration where the flow is kept constant, with variable SpO₂ values. In preterm infants receiving mechanical ventilation, automated O₂ control results in more time spent within the intended SpO₂ target.^11–13 In a healthy adult model with induced hypoxemia, such a system was more efficient to maintain SpO₂ within the oxygenation target, while ensuring a significant reduction of hypoxemia and hyperoxia periods, as compared with constant O₂ flows.¹⁴ Its efficacy has also been validated in hospitalised patients with COPD¹⁵ or during the early emergency care of patients with acute respiratory distress.¹⁰

Following major abdominal or thoracic surgery, the risk of hypoxemia may be high while considering patients’ clinical status (obstructive sleep apnoea (OSA), restrictive pathologies related to obesity, frequent comorbidities), the type of surgery and anaesthesia.^16–21 Hypoxemia may occur either during the immediate postoperative period (they are mainly related to surgery or anaesthesia) or may be delayed up to 3 days without clear trigger or underlying pathologies. The potential interest of the FreeO₂ system, using artificial intelligence closed-loop adjustments and predictive analytics, will thus be (1) to perform frequent and rapid O₂ adjustments in response to oxygenation condition variations (up to each second), or to any physiological condition changes (movement, speech, eating, toilet); (2) to enable remote monitoring and data recording in isolated clinical settings (ie, non-ICU surgical ward), in order to detect clinical deterioration at a very early stage through integration and fusion of informations; (3) to avoid maintenance of unnecessary high O₂ flow that may be potentially deleterious (hypercapnia worsening, coronary and/or cerebral artery vasoconstriction, etc).

Objectives

The aim of the study is to assess the use feasibility of the FreeO₂ system so as to deliver oxygen automatically in the postanaesthesia care unit in a patient population admitted for major abdominal and thoracic surgery. Our hypothesis is that the FreeO₂ system will provide a better control of the oxygen saturation and reduce postoperative hypoxemia.

Trial design

The FreeO₂ PostOp study is an investigator-initiated, prospective, multicentre, randomised, controlled open trial on medical devices comparing two strategies of oxygen therapy following major surgeries with either standard treatment or automated closed-loop oxygenation. Patients are randomly assigned to the Standard (manual O₂ administration) or FreeO₂ group (automated closed-loop O₂ administration).

Methods: patients, interventions and outcomes

Study setting

The FreeO₂ PostOp study is taking place in five different university hospitals in France and Canada (Brest, Clermont-Ferrand, Montpellier, Poitiers, Quebec). The entire study will be performed in accordance with the International Conference on Harmonisation and Good Clinical Practice guidelines.²²

Eligibility criteria

Inclusion criteria

To be included, adult patients (≥18 years) may be screened for scheduled abdominal or thoracic surgery during the anaesthesia consultation and be considered as requiring general anaesthesia with an expected duration of 2 hours or more in the participating centres.

Patients with intermediate to high risk for postoperative pulmonary complications following abdominal surgery with general anaesthesia are eligible for participation. To identify such patients, the ARISCAT risk score will be used.²³ An ARISCAT risk score ≥26 is associated with an intermediate to high risk for postoperative pulmonary complications.

Informed consent is to be signed before the surgery (figure 1).

Figure 1.

Participant timeline of the FreeO₂ PostOp study. Inclusion is performed during the anaesthesia consultation, prior to a scheduled major surgery (≥2 hours’ duration; ARISCAT score ≥26). Eligibility criteria are verified and the patient needs to sign the informed consent. Randomisation is performed no later than 1 hour following extubation in the recovery room. Clinical data will be recorded each hour during the first 3 hours of care and twice daily for up to 3 days. Continuous SpO₂ recording will be performed in each randomisation groups for up to 3 days. Outcome parameters will be assessed at day 28. LOS, length of stay (days); SpO₂, pulse oximetry (%).

Exclusion criteria

Patients fulfilling one or more of the following criteria will not be included: life-threatening condition requiring an unplanned emergent surgery, lack of informed consent prior to randomisation, non-adult patients (age <18 years), patients with a Body Mass Index ≥35 kg/m², patients with OSA, pregnant or lactating women, perturbed or non-cooperative patients, to provide a relatively homogeneous study population and avoid potential confounding factors in the interpretation.

Randomisation criteria

Patients will be randomised if they fulfil all the following criteria: availability of the FreeO₂ prototype; absence of criteria of severity justifying the immediate use of ventilatory support (loss of consciousness with a Glasgow Coma Score ≤12), serious ventricular rhythm disorders, haemodynamic instability (systolic blood pressure <80 mm Hg or recourse to vasopressors), cardiac or respiratory arrest, pH <7.35 and PaCO₂ >55 mm Hg (if measured), and necessity of an O₂ flow less than 15 L/min to maintain a SpO₂ higher than 92%; no emergent surgery required for an adverse event; pulse oximetry signal is available.

Study intervention

Patients eligible for randomisation will be randomly assigned to the Standard or FreeO₂ group (figure 2). The maximal study duration is 3 days. The study was stopped before 3 days if patient was discharged earlier from hospital or in case of discomfort while keeping nasal prongs and sensors.

Figure 2.

Consort diagram of the FreeO₂ PostOp study. Randomisation will be performed after verification of the eligibility criteria and it will be stratified according to the presence of a medical history of chronic obstructive pulmonary disease (COPD). For patients without COPD, the target will be a SpO₂ range=92%–96%; for patients with COPD, the target will be a SpO₂range=88%–92%. Patients will be assigned either to standard continuous O₂ administration (Standard: manual adjustment) or automated closed-loop O₂ administration (FreeO₂: automated adjustment, up to each second); in both groups, continuous SpO₂ recordings will be performed during up to 3 days according to the FreeO₂ system. SpO_2, pulse oximetry (%).

Automated O₂ administration is performed using the FreeO₂ system (Oxynov) that is set to maintain SpO₂ between 92% and 96% for patients without COPD or between 88% and 92% for patients with COPD. FreeO₂ is equipped with a SpO₂ monitor and an electronically controlled valve that automatically adjusts O₂ flows from 0 to 20 L/min on a per-second basis, with a 0.1 L/min precision, according to a closed-loop algorithm in order to reach the predetermined SpO₂ target.¹⁴ Conventional O₂ is administered using manual flowmeters, according to standard procedures. All participating units were encouraged to use the same standardised SpO₂ target as in the automated O₂ administration group, as recommended in international guidelines.^{1 2} In both arms, oxygen can be administered either using nasal prongs for low flow (O₂<6 L/min) or standard face mask in all cases (O₂=0–20 L/min).

Continuous oximetry recordings are performed in each group during the 3 days of the study, using the FreeO₂ monitoring system connected to Nonin 6000 CA flexible adult single-use digital sensors. Position of the sensors is to be controlled at least every 12 hours. In the FreeO₂ group, the FreeO₂ will be used both for oxygen administration and for recording. In the Standard group, the FreeO₂ will be used in the recording only and the oxygen administration will be set manually.

Standard procedures

Considering the variety of pathological cases for patients attending the recovery room, medical treatment, including the other respiratory support, is determined by the attending physicians based on clinical needs assessment. All other aspects of patient care after inclusion in the study, including fluid administration, prophylactic antibiotics and postoperative pain management, are made at clinicians’ discretion based on the expertise of the staff at each centre and routine clinical practice.

Primary outcome measure

The primary outcome measure will be the percentage of time spent in the target zone of oxygen saturation, during a 3-day time frame. In both groups (Standard and FreeO₂), patients will be connected to the FreeO₂ system to enable continuous oximetry recordings of one SpO₂ value per second, either in the recording mode for patients assigned to the Standard group or in the automated closed-loop mode for patients assigned to the FreeO₂ group.

The target zone of oxygen saturation is SpO₂=88%–92% for patients with COPD and 92%–96% for patients without COPD.

Secondary outcome measures

Several secondary outcome measures will be evaluated during a 3-day time frame: nursing workload assessed by the number of manual O₂ flow adjustments and airway management procedures (twice daily assessment); time spent with severe desaturation (SpO₂ <85%); time spent with hyperoxia (SpO₂ >98%); VO₂ measured at the end of administration. All data related to oxygenation will be recorded in both groups using the FreeO₂ device, enabling us to qualify desaturation using the Oxygen Desaturation Index at different levels (ODI2% and ODI4%).

Otheroutcome measures will be assessed during a maximal 28-day time frame: duration of oxygen administration during hospitalisation, number of complications related to the administration of oxygen; frequency of use of ventilation (invasive or non-invasive); duration of hospitalisation; survival rate.

Participant timeline

Figure 1 shows the participant timeline of the FreeO₂ PostOp study.

Sample size

Based on previous studies and data from the literature,^{10 20} we estimated an 85% time within oxygenation range with the automated closed-loop oxygen administration system (FreeO₂) during the 3 days following surgery and a SD equal or more than 30%. A total number of 180 patients will be needed to demonstrate a 15% decrease in absolute difference between the Standard and FreeO₂ groups (90 patients in each group). The α risk is 5% and the β risk is 10% with a bilateral formulation. The expected duration of patient enrolment will be 3 years.

Methods: assignment of intervention

Allocation and sequence generation

A computer-generated randomisation is performed in the postrecovery room within a 1-hour delay after endotracheal extubation. It is performed using random blocks in a 1:1 ratio, with the use of a centralised web-based management system (Clinfile). Stratification is performed either according to the study centre and a medical history of COPD. After randomisation, treatment is to be initiated within less than 1 hour.

Blinding

Although the individual study assignments of the patients will not be masked, the coordinating centre and all the investigators will remain unaware of the study group outcomes until the data will be locked.

Methods: data collection, management and analysis

Data collection and management

At the time of enrolment, physiological characteristics, coexisting medical conditions, surgery and anaesthetics characteristics, and oxygen flow rates are recorded. In both groups, pulse oximetry, respiratory and heart rate are continuously monitored using a dedicated software enabling data extraction from the FreeO₂ device. Each hour during the first 3 hours following randomisation and daily during the following 3 days, all standard clinical parameters are collected. Evolution and clinical outcomes are monitored at day 28.

Data are collected and recorded on an electronic case-report form (e-CRF) by a trained investigator or research assistant at each centre. A blank copy of the e-CRF can be printed from the e-CRF; this enables the investigator or research assistant to fill it out with the data of the included patients, which will be captured. Once data collection has been completed, the investigator or research assistant shall sign and date the copy. This document will constitute an integral part of the patient’s medical records; as such, it shall be retained permanently. Data recorded in the e-CRF that originate in source documents must be consistent with each other; if they are not, the differences have got to be justified and documented.

Statistical methods

All the analyses will be performed by the study statistician.

A predefined statistical analysis plan will be followed. The analysis will be performed on an intention-to-treat basis after validation by a blind review committee of the inclusion and exclusion criteria for each patient. All the analyses will be conducted by the biostatistics department of the Centre d’Investigation Clinique from Brest University Hospital (CIC INSERM 1412) using SAS V.9.3 statistical software (SAS Institute, Cary, North Carolina, USA). A two-tailed p value equal or less than 0.05 will be considered as statistically significant.

Descriptive analysis of patient groups at baseline

The continuous variables will be analysed using standard parameters (median, IQRs and extreme values, or mean and SD), while indicating the number of missing data. The category variables will be presented in the form of absolute frequency and percentage, in each treatment group.

The criteria of subject selection will be verified based on the data recorded in the electronic case reports. Falsely included subjects such as those lost to follow-up will be described. Deviations from the protocol will be described and analysed on a case-by-case basis.

Analysis pertaining to the main criteria of evaluation

The percentages of time within the considered SpO₂ range will be compared between the two groups by means of a variance analysis, according to stratification.

Analysis pertaining to the secondary criteria of evaluation

Secondary criteria of evaluation will be compared between the two treatment groups by means of Student’s t-test (or the Mann-Whitney U test, if necessary) for continuous quantitative variables and by means of the χ² test (or Fisher’s exact test) for qualitative variables.

Methods: monitoring

Data monitoring

Before starting patient enrolment, all physicians and other healthcare workers in the ICU attend formal training sessions on the study protocol and data collection.

An investigator at each centre is responsible for daily patient screening, enrolling patients in the study, ensuring adherence to the protocol and completing the e-CRF. Research assistants regularly monitor all the centres on site to check adherence to the protocol and the accuracy of the data recorded.

Harms

The investigator is allowed to temporarily or permanently discontinue the participation of a patient in the study for any reason that would optimally serve the interests of the subject, particularly in case of serious adverse events suspected to be associated with the type of oxygenation method that is used.

Auditing

The trial is overseen by a steering committee and an independent safety monitoring board composed of three independent experts (Jean-Pierre Frat, Christophe Guitton, Alain Mercat). All centres are monitored by the promoter to check adherence to the protocol and accuracy of recorded data. An investigator at each centre is responsible for enrolling patients and ensuring adherence to the protocol. Research assistants at each centre are responsible for patients’ follow-up and for completing the e-CRF.

Ethics and dissemination

The French national agency for drug and biomedical devices security (ANSM) approved the use of the FreeO₂ device within this study on 5 August 2014. The main modification of the protocol was made after obtaining CE marking for the device (protocol version no. 3); other amendments were related to investigators’ changes and an extension of the study length (final version: no. 5). The study sponsor notified all amendments to investigators.

Consent or assent

The patient is included after having provided a written informed consent to the investigator according to the decision of the central ethics committee. If the patient is not able to understand the information given, he/she can be included if the same procedure is completed with a next of kin. After the patient’s recovery, he/she will be asked if he/she agrees to continue the trial.

Patient and public involvement

Patients and the public were not involved in the design or conduct of the study. There is no plan to disseminate the results to study participants.

Confidentiality

Data will be handled according to French law. All original records will be archived at trial sites for 15 years. The clean database file will be anonymised and kept for 15 years.

Dissemination policy

The protocol is reported according to the SPIRIT guidelines.²⁴ Findings will be submitted to peer-reviewed journals and presented at local, national and international meetings and conferences according to CONSORT guidelines,²⁵ to publicise and explain the research to clinicians, commissioners and service users.

Authorship eligibility guidelines will take into account participation to the protocol design and writing, final analysis of the results, patients’ recruitment and final manuscript revision.

Discussion

In postoperative patients, pulmonary function is markedly altered both by general anaesthesia, mechanical ventilation and surgery. Postoperative respiratory complications following surgery are the second most frequent complications after surgery¹⁹ and considered as a major cause of morbidity and mortality.^{17 26 27} The FreeO₂ PostOp trial is the first clinical evaluation of automated oxygen titration in patients undergoing major abdominal or thoracic surgeries. The objectives of the study are to compare the oxygenation parameters (time in the oxygenation target, time with hypoxemia and with hyperoxia) with usual oxygen therapy and automated oxygen titration and weaning.

Risk and consequences of hypoxemia

There is a good level of evidence and good acceptation by the medical community that hypoxemia is harmful,²⁸ especially in adult patients with myocardial ischaemia²⁹ and neurotrauma.³⁰ Data also suggest that even short periods of hypoxemia may promote significant negative haemodynamic effects.³¹ In an animal model, right ventricular dilation was observed with only 2 hours of daily hypoxemia.³²

During the postoperative period, the occurrence of marked oxygen desaturation related to periodic apnoea and hypoventilation for several hours has been recognised for a long time,^{33 34} with potentially severe consequences such as myocardial ischaemia.³⁵ The incidence of hypoxemia is high in the recovery room (10%–50%),²⁰ and up to 50% of postoperative patients will demonstrate episodic hypoxemia in the absence of O₂ therapy.^{36 37} It has also been shown that desaturation episodes are more frequent on the first night after surgery, but may also worsen 3–5 days postoperatively, especially in patients with OSA.^38–40

Pathophysiology of these O₂ desaturations is complex but may be due either to the patient’s condition itself (advanced age, COPD, diabetes, obesity, etc), respiratory mechanics modifications (reduction of functional residual capacity, atelectasis, thoracoabdominal compliance decrease), but also to the use of pharmaceutical agents that are given during surgery (anaesthetics and neuromuscular blocking drugs) or those that are given to relief postoperative pain (opioids and sedatives).³⁴

Our hypothesis is that desaturations will be reduced with the automated oxygen titration (FreeO₂ group), as the oxygen flow is titrated every second and that desaturations will immediately lead to increasing of oxygen flowrate. In previous studies comparing manual and automated oxygen administration, it was demonstrated that time with hypoxemia was reduced with FreeO₂.^{10 15} However, no data are available during the postoperative care.

Automated weaning of oxygen

Oxygen supplementation has initially been promoted to decrease postoperative hypoxemia, even if it is also known that it will not have any effect on the overall number of central or obstructive apnoea, neither atelectasis.³³ Not all patients will benefit from systematic O₂ administration,⁴¹ but probably only specific patients with a high-risk profile, especially those following long duration and major thoracic or abdominal surgery.^42–44 Moreover, standard continuous O₂ therapy tends to reduce but not abolish the occurrence of desaturation,³⁶ given the fact that only point checks are made to adjust O₂ flow to actual patients’ needs.

It has also been demonstrated that the duration of O₂ therapy was an independent risk factor for developing postoperative respiratory complications. Patients who require O₂ for ≥75% of recovery room time (or greater than 90 min) appear to be at greater risk of developing respiratory complications.⁴⁵ This fact may suggest that some patients are not adequately screened for risk factors such as OSA by standard preanaesthesia testing, and that a device dedicated to continuous monitoring of O₂ administration (either alarms on duration and flow variations) may help to detect such high-risk patients. The increased resource use in patients with longer oxygen therapy requirement in the recovery room likely reflects the increase in occurrence of pulmonary respiratory complications requiring invasive and non-invasive ventilatory support, especially on the day of surgery.

In L’her et al’s study, it was shown that the partial or complete oxygen weaning was significantly increased with automated oxygen titration in comparison with standard oxygen administration, in patients managed in the emergency department during 3 hours.¹⁰

In the specific setting of postoperative patients, the reduction of the weaning time may improve the efficiency of the turnover of the patients in the recovery room.

Risk and consequences of hyperoxemia

Few studies have promoted the use of hyperoxia during and after colorectal surgery to reduce wound infection,⁴⁶ but this is not recommended in routine, given controversial data.^47–49 In a post hoc analysis of the PROXI trial, the authors even pointed out the potential risks of acute coronary syndromes associated with perioperative hyperoxia.⁵⁰ The pathophysiological risks associated with hyperoxia are described for a long time, especially in patients with COPD.⁴ The first recommendation to adjust oxygen flow rates in order to reduce the risks of hyperoxia was published in the early 1960s,⁵¹ and several more recent guidelines have reiterated similar recommendations.^{1 2} The recent demonstration in a large randomised controlled trial (RCT) of an increased mortality for ICU patients assigned to a standard O₂ therapy practice, as compared with a more conservative one (absolute risk reduction 0.086 (95% CI 0.017 to 0.150); p=0.01), has revived the debate about potential harm of excessive oxygen therapy in an unselected patients’ population.⁵² Adverse effects of hyperoxia could be mediated through a higher oxidative stress, but also increased coronary^{6 7} and cerebral artery resistances,⁸ all being associated with a potential clinical impact.^{53 54} Such potential adverse events related to hyperoxia clearly mandates attention in avoiding unnecessary O₂ administration. It was demonstrated in two randomised trials from our team^{10 15} that time with hyperoxia could be reduced with automated oxygen titration, and the hypothesis of the present study is that this problem will also be reduced after thoracic or abdominal surgeries.

Clinical data with automated oxygen titration

Several systems have been developed to titrate oxygen flow rate in neonates and in adult patients.^{10 14} In a previous RCT on adult patients admitted to the ED for acute hypoxemic respiratory failure, the use of automated O₂ administration was found superior to manual O₂ administration to improve the time spent within oxygenation targets, with a between-group difference of 29%,¹⁰ as already observed in other studies on O₂ automated administration.^{55 56} In L’Her et al’s study, patients experienced less time with hypoxemia and hyperoxia in the FreeO₂ group.¹⁰ When receiving automated oxygen, partial or complete oxygen weaning was more frequent during initial care, as compared with standard manual O₂ administration.

As automated closed-loop O₂ administration is not the standard of care for postoperative patients, we did not design this study as a non-inferiority study but rather as a superiority study. One limitation of the study is that investigators are aware of the inclusion group, and blinding is difficult in studies with respiratory support. However, given the characteristics of the two strategies under evaluation, a double-blind trial was not possible. Second, the assessment of the oxygenation status could be considered as more precise by analysing blood gas sample rather than SpO₂. However, this would not enable continuous oxygenation monitoring up to 3 days. Moreover, only continuous of non-averaged SpO₂ values do enable precise and rapid adjustments of the O₂ flow, in response to exact patients’ needs.

Noteworthy, all patients in both groups will be continuously monitored using the same oximeter, which may represent a strength of this study; therefore, the FreeO₂ PostOp will also represent the largest prospective study comparing two oxygenation strategies over such a period in the postoperative setting.

In conclusion, the FreeO₂ PostOp trial is a pragmatic RCT designed to test the hypothesis that automated closed-loop O₂ administration is superior to standard manual O₂ administration, during the postoperative care of patients with major abdominal or thoracic surgery. To the best of our knowledge, the FreeO₂ PostOp trial is the first to evaluate the usefulness of automated closed-loop O₂ administration after extubation in such an indication.