Evaluation of awake prone positioning effectiveness in moderate to severe COVID-19 [version 1; peer review: awaiting peer review]

Evidence mainly from high income countries suggests that lying in the prone position may be beneficial in


Introduction
Placing patients with respiratory failure in the prone position is associated with benefits in respiratory mechanics and blood oxygenation 1 . In mechanically ventilated patients with acute respiratory distress syndrome (ARDS), prone positioning is associated with improved survival 2 . The huge burden of respiratory failure resulting from COVID-19 has stimulated interest in understanding prone position in those receiving no or lesser degrees of respiratory support -often termed 'awake prone position' (APP).
To date, several randomised controlled trials of APP in COVID-19 have been performed. Since this protocol was written, 10 of these have been included in a meta-analysis 3 . The trials included varied in size (median 60 participants); location (8/10 exclusively high income countries) and inclusion criteria (room air, supplemental oxygen, non-invasive ventilation). Almost all participants were recruited in the early phase of the pandemic in unvaccinated populations. Overall the metaanalysis concluded APP was beneficial, and APP has been incorporated into many guidelines for treatment of COVID-19 in many countries, including low-and middle-income countries (LMICs). Similar to the effect observed in mechanically ventilated patients with ARDS, success of APP was noted in patient groups with the longest duration of prone positioning as well as those in ICU environments [3][4][5] . Relationship with disease severity is less clear, although the largest study (Ehrmann et al's) recruited only those with HFNC 4 . A further nonrandomized controlled trial of 500 patients admitted to 2 centres in North America has since been published indicating no substantial difference of APP at 5 days with a suggestion of possible harm at 14 and 28 days 6 .
Much of the interest in APP for moderate to severe COVID-19 disease, is related to its perceived low-cost and simplicity of implementation. In reality, even in high income countries, effective implementation is challenging, especially when carried out under pandemic caseloads. In one randomized controlled trial in the USA, where trained admitting teams instructed patients in the APP protocol, only 37% of patients in the intervention group actually attempted to lie prone 7 . In only a few of the reported trials were patients supervised (and if necessary helped) into the prone position. In resource-limited settings, where there are already restricted numbers of healthcare staff, this additional workload is a significant demand on the healthcare service, and likely requires additional staffing, increasing the overall cost to the healthcare system. The only two trials of APP from LMICs were both carried in ICU settings where there are likely to be significantly more staff on hand: in India, investigators succeeded in 43% of the intervention group spending ≥6 hours/day prone 8 , and in Mexico median 8.6 hours a day 4 . These issues of lack of staff and limited duration were cited as reasons for lack of efficacy of the recently published multicentre COVID-PRONE study and an accompanying editorial emphasised that future trials need to focus on optimal means of sustaining APP 9 . With widespread vaccination, many patients requiring respiratory support for COVID-19 are more likely to be elderly, frail and with significant comorbidity. The feasibility of these patients achieving prolonged periods in the prone position without significant help is unknown.
For these reasons, we are conducting this randomized controlled trial of APP in COVID-19 in Vietnam. Importantly it employs a dedicated team and wearable technologies to monitor patients and accurately quantify changes in vital signs and patient position, thus attempting to ensure maximum durations of APP, but also utilize new technologies for monitoring which may ultimately reduce staff time performing these tasks. Data capture uses a combination of special case report form but also utilizes an already functioning electronic data registry where routine clinical information is recorded by a dedicated data-entry team 10 . Whilst this trial was designed at the height of the Delta wave pandemic in Vietnam, the situation has evolved to currently Omicron variant and widespread vaccination coverage. As a result, unlike most other preceding studies of APP, this trial will enrol a different patient population.

Background and scientific rationale
Supplemental oxygen is recommended to maintain oxygen saturations in those with COVID-19 associated acute respiratory failure. Many patients require escalation of therapy, from simple low-flow systems to higher flow methods, non-invasive ventilation or endotracheal intubation and invasive mechanical ventilation 11 . Escalation of therapy necessitates increased utilization of healthcare resources such as oxygen, equipment and skilled staff. Whilst already in short supply in resource-limited countries, these are even further limited in the current pandemic situation 11 .
In patients with severe acute respiratory distress syndrome (ARDS) receiving mechanical ventilation, prone position has been shown to increase survival and respiratory outcomes 1 . Notably the PROSEVA trial, where patients with severe ARDS (PaO 2 :FiO 2 ratio <150 mmHg) placed prone for a median 17 hours a day resulted in a reduced risk of death at 28 and 90 days compared to supine position (hazard ratio of death 0.42; 95%CI 0.26-0.66) 2 . Furthermore, patients placed in the prone position had shorter duration of ventilation and more successful extubation events than those in the supine position.
Prone positioning has a variety of effects that may have beneficial effects on pulmonary and cardiovascular physiology and ventilation/perfusion matching. These include reducing pulmonary compression by abdominal or mediastinal organs; reducing antero-posterior pulmonary pressure gradients and localized hyperinflation; improving right heart function and reducing dead space. Improved oxygenation and carbon dioxide clearance may allow reduction in FiO 2 or ventilator settings, reducing iatrogenic pulmonary and systemic injury 1 .
Currently many guidelines recommend prone position in patients with severe COVID-19 ARDS requiring mechanical ventilation 12 . However, placing a patient in a prone position is a complicated process and, in severely ill patients, requires a team of trained operators and considerable time. In resource-stretched units, providing these safely is challenging. Furthermore, prone positioning may be associated with increase in events such as pressure sores, nerve compression injury or accidental displacement of lines or endotracheal tubes 13 . Evidence suggests that prone position may increase the work of breathing, thus although oxygenation improves, this benefit may be offset by fatigue 14 .
Prone positioning may be easier to achieve in awake, self-ventilating patients. Data from observational studies has indicated that prone positioning in such patients ('awake prone positioning') is associated with improved oxygenation, reduced intubation and mortality in a Low and Middle Income Country (LMIC) setting 15 . A Delphi consensus of 39 experts from 20 countries reported >90% agreed that awake self-proning could improve oxygenation in COVID-19 acute respiratory failure 16 . However only 54% agreed, based on their own experience, that the procedure may reduce the need for mechanical ventilation.
A recent 'meta trial' combining data from five randomized controlled trials compared prone positioning with supine position in 1126 patients receiving oxygen via high flow nasal canulae (HFNC) for COVID-19 pneumonia. Treatment failure, defined as intubation or death, occurred in 40% of 564 patients assigned to awake prone positioning and 46% of 557 patients assigned to standard care (relative risk 0·86 (95% confidence interval 0·75, 0·98). The study showed a reduction in requirement for intubation in those treated with prone position (hazard ratio 0.75, 95% confidence interval 0.62, 0.91) and a non-significant reduction in mortality (hazard ratio 0.87, 95% confidence interval 0.68, 1.11) 4 . Whilst the majority of studies and participants were from high-income countries, approximately 400 enrolled patients were from Mexico. Of note in this study, patients in the prone group achieved a median 5 hours a day in the prone position. Evidence from patients with ARDS receiving mechanical ventilation indicates that longer periods in the prone positon are more beneficial 2 . An observational study of patients receiving supplemental oxygen by mask for COVID-19 respiratory failure in Brazil reported a relative risk of mechanical ventilation of approximately 0.4 in those treated with prolonged periods of prone position of approximately 12 hours duration daily 15 . Therefore it is possible that longer periods of prone positioning are associated with better outcomes.
In summary, there are convincing data supporting the use of prone position in reducing mortality and ventilation time in patients with ARDS receiving mechanical ventilation. Data also support the use of awake prone position to improve oxygenation in patients with COVID-19 related acute respiratory failure, and randomized control trial data suggest this is may also prevent intubation. However, it is unclear whether patients on lower degrees of respiratory support (eg oxygen via facemask or nasal canulae) also benefit, nor whether data from well-resourced healthcare environments apply in LMICs such as Vietnam, particularly under pandemic situations when considerable resources are required to help patients maintain prone positioning.

Primary objective
To determine whether prone positioning of hospitalized Vietnamese patients with moderate to severe COVID-19 for ≥ eight hours a day reduces the need for escalation of respiratory therapy compared to standard care.

Secondary objectives
1. To determine whether prone positioning with a protocol aiming for ≥ eight hours a day results in reduced intubation rates, improved mortality and shorter duration of hospital stay compared to standard care.
2. To compare changes in FiO 2 , SpO 2 , respiratory rate and heart rate that occur during prone position of hospitalized Vietnamese patients with moderate to severe COVID-19.
3. To determine whether prone positioning for ≥ 8 hours a day is associated with reduced oxygen utilization compared to standard care.

4.
To determine safety of prone positioning.

Study design and endpoints
Study design Pragmatic open label randomised controlled trial

Study population
All adult patients (≥18 years old) presenting to the study sites due to probable or proven COVID-19 pneumonia, subject to the inclusion and exclusion criteria.

Intervention
Standard care: will consist of routine clinical care, including any advice to lie in prone position as routinely recommended by participating sites. To reduce bias, the study team will make visits at a similar schedule to those to patients in the intervention group, however these visits will be confined to general advice and measurement of vital signs. Patients will receive written advice more general in nature about COVID-19 disease.

Intervention.
A special intervention team will visit patients' rooms aiming for patients to maintain the prone position for at least 8 hours a day. The team will give written and verbal advice and if necessary aid patients' positioning themselves in the prone position. The exact duration of prone sessions will be determined according to ward schedules to take account of nursing procedures, meal times and mitigation strategies (see Extended data). Other methods to encourage the maintenance of prone position includes phone calls to patients, carers or education of carers.
Compliance with the intervention will be evaluated by observation (manual and using in-room cameras) at fixed time intervals. In a subgroup of 100 patients accelerometer/gyrometer devices will be used to measure movement and position (this will be expanded to all patients if equipment is available).

Study endpoints Primary endpoint
Escalation of respiratory therapy within 28 days of randomization, defined as any of the following: • Escalation to next level respiratory support (with lowest level nasal canulae or face mask, escalating through HFNC to NIV or mechanical ventilation).

Secondary endpoints
• Requirement for intubation and mechanical ventilation within 28 days of randomization • Duration of prone position Additional endpoints, measured in this group of patients but reported separately, include acceptability from patients' perspectives and hospital direct medical costs.

Sample size
Our sample size is based on local data showing a current treatment failure rate of approximately 52%, a relative risk of treatment failure estimated at 0.8 for the intervention, corresponding to an absolute risk reduction from 52% to 40 %. To detect this reduction with 80% power at the two-sided 5% significance level, 300 patients are required in each arm, giving a total sample size of 600 allowing for exclusions, loss to follow up and withdrawals. (also see Extended data 4)

Participant withdrawal of participation
Each participant has the right to withdraw from the trial at any time. In addition, the Investigator may discontinue a participant from the trial at any time if the Investigator considers it necessary for any reason including: • Ineligibility (either arising during the trial or retrospectively having been overlooked at screening) • An adverse event which requires discontinuation or inability to continue to comply with trial procedures

• Withdrawal of Consent
Handling of participant withdrawal or termination If a patient or the representative, who has given consent on their behalf, chooses to discontinue trial treatment, they should be followed up (providing they are willing) and encouraged to follow the study procedures in lieu of withdrawing from the trial. If they do not wish to remain on trial follow-up, however, their decision will be respected and the patient will be withdrawn from the trial completely. This will be recorded on the OUTCOME CRF. The reason for the patient withdrawing should be ascertained wherever possible. If a participant withdraws from the trial, they may agree that the medical data collected during their previous consented participation in the trial will be kept and used in analysis, or if they do not consent to this, all data will be excluded from analysis.
Participants may change their minds about stopping trial follow-up at any time and re-consent to participation in the trial. Participants who stop trial follow-up early will not be replaced, as the total sample size includes adjustment for losses to follow-up.

Study procedures
Informed consent Informed consent to enter into the trial and be randomized must be obtained from all participants. If the patient lacks capacity to give consent due to the severity of their medical condition (eg acute respiratory failure), then consent may be obtained from a relative acting as the patient's legally designated representative or -if a suitable relative is not available after reasonable efforts to locate one -an independent doctor. Further consent will then be sought from the patient if they recover sufficiently.
Individuals trained and responsible for taking consent will be documented on the trial's Delegation Log (with signatures). This should be, if appropriate, after explanation of the aims, methods, benefits and potential hazards of the trial and before any trial-specific procedures are performed, including for the screening assessment.
It must be made completely and unambiguously clear that the participant (or their relative) is free to refuse to participate in or withdraw from all or any aspect of the trial, at any time and for any reason, without incurring any penalty or affecting their subsequent treatment. This will be stated explicitly in the participant information sheet. If consent was provided by a relative, the participant should be consulted and consent recorded if and when they have the capacity to do so.
Copies of consent forms must be kept in the investigator site file and a copy given to the participant or family. Considering the nature of the diseases, the study patient may be treated in an isolated area, where no paper document is allowed to be moved out of the infectious area. The study will consider applying electronic ICF capture on tablet or any other appropriate method for obtaining valid ICF.

Randomization
Patients will be randomly allocated 1:1 to either standard care or intervention, stratified according to study ward. A stratified, computer-generated randomization list will be created using block randomization with variable block length and incorporated into secure internet accessible randomization software. Once a consented patient develops eligibility criteria, or an eligible patient is consented, the initials and date of birth of the patient will be entered into the software by study staff. Based on the randomization list, the software will produce the treatment allocation, which will be displayed and recorded in the study database. All entries and outputs of the software will be auditable.

Clinical monitoring
Study staff will make regular visits to all patients in the study, to carry out study procedures, assess compliance with prone positioning and, if necessary, record vital sign data (see Extended data). Additionally, we will use in-room cameras and wearable accelerometer / gyrometer devices to verify prone positioning of participants. Using these methods, we aim to evaluate compliance with prone position hourly for eight hours a day.
We aim to monitor all patients in this study using continuous vital sign monitors to measure vital signs hourly. For patients with appropriate monitoring in situ, continuous vital sign waveform data will be recorded. For others, intermittent vital sign summary data will be recorded with a frequency at least of routine clinical monitoring schedules.
Supplemental oxygen flow rates and method of delivery will be recorded daily to allow calculation of oxygen utilization.
Acceptability of the intervention from patient perspective will be evaluated whenever patient is recovery and available for telephone call using a Likert scale to assess comfort and acceptability (see Extended data 10).
Procedures for assessing safety An independent data safety and monitoring committee (DSMB) will oversee the safety of the trial participants. For details on the DSMB and schedule of assessment see later section.
Study staff will perform daily monitoring for safety events. It is recognised that this study involves seriously ill patients in whom disease progression and adverse events are common.
In view of this monitoring will particularly focus on adverse events related to the study intervention and serious adverse events as detailed below.

Adverse events
An adverse event (AE) is defined as any unfavourable and unintended sign (including an abnormal laboratory finding), symptom, or disease temporally associated with the use of the medical treatment or procedure that may or may not be considered related to the medical treatment or procedure. The severity of all AEs in this trial should be graded in line with the toxicity gradings in Toxicity grading and management (CTCAE).
Adverse events will be graded for severity: When an AE occurs, the investigator responsible for the care of the participant must first assess whether or not the event is serious using the definition below.
A Serious Adverse Event (SAE), is any AE that: • Results in death, • Is life-threatening, • Requires hospitalisation or prolonged or existing hospitalisation, • Results in persistent or significant disability or incapacity, or • Consists of a congenital anomaly or birth defect, or Causality. The investigator must assess the causality of all serious events or reactions in relation to the trial therapy (prone positioning) using the definitions below. Reporting of adverse events Serious adverse events definitely related to the intervention and unexpected serious adverse events will be reported to OUCRU CTU within 24 hours after the investigatory/ study team become aware of the event. Investigators should notify the OUCRU CTU of these predefined SAEs occurring from the time of randomization until the participant finishes their follow-up. CTU will perform an initial check of the report, request any additional information. The causality assessment given by the local investigator at the hospital cannot be overruled; in the case of disagreement, both opinions will be provided in any subsequent reports. All SAE information must be recorded on an SAE form and sent or emailed, to CTU. Additional and further requested information (follow-up or corrections to the original case) will be detailed on a new SAE Report Form and sent to CTU.
The OUCRU CTU is undertaking the duties of trial sponsor with regard to safety reporting and is responsible for the reporting to the regulatory authorities and the research ethics committees, as appropriate.
The SAEs will be reported as soon as possible to the site ethics committee (EC). An initial written report of those resulting in death, or that are life threatening, has to be reported urgently within seven working days of the study team becoming aware of the SAE. Others must be reported within 15 working days of the study team becoming aware of the SAE. Additional medical information of the SAE's development must be reported in an additional report until the trial subject recovers or stabilises without further changes expected. The format and content of the initial report should follow the relevant Ethics Committee report template and include all information available at the time of reporting.
All specified SAEs will be reported to OxTREC in the annual review form and to the DSMB in accordance to the DSMB charter.

Study halting rules
An independent DSMB will oversee the trial. Serious adverse events will be reported to the DSMB within 10 days of occurrence and followed-up until resolution. The DSMB will perform a safety analysis after the first 60 patients have completed 28-day follow-up or died. Stopping for harm will be considered if a safety issue emerges which is sufficiently large, in the judgement of the DSMB, to suggest that continued exposure of patients to the intervention is unethical. The DSMB will be able to mandate additional safety analyses at any time-point they deem fit and will determine the schedule for further analyses throughout the trial.
At the interim analyses, the DSMB will receive a report including summaries of mortality, serious adverse events, by treatment arm. The report will be prepared by the DSMB statistician and distributed to all DSMB members for review. Based on these data, the committee will make recommendations on the continuation, cessation or amendment of the study. The study statistician will aid in setting-up the code for generating the interim analysis summaries.
As the dissemination of preliminary summary data could influence the further conduct of the trial and introduce bias, access to interim data and results will be confidential and strictly limited to the involved independent statistician and the monitoring board and results (except for the recommendation) will not be communicated to the outside and/or clinical investigators involved in the trial.
Further reviews will be at the discretion of the DSMB. All DSMB reports, replies or decisions will be sent to the responsible Research Ethical Committees.

Statistical considerations
Statistical and analytical plans Study analysis and presentation of results will be according to an a priori defined statistical analysis plan which will be completed before database locking and based on the considerations below.
The primary analysis population for all analysis is the full analysis population containing all randomized patients. Patients will be analysed according to their randomized arm (intentionto-treat). Analyses for the primary endpoint will be repeated on the per protocol population which excludes the following patients: patients not receiving the randomized intervention and other major violations of inclusion/exclusion criteria or study procedures.
The primary outcome measure, will be summarized as x/n (%) in each group and compared between the groups based on a logistic regression model with the intervention as the main covariate. Time-to-event analysis will be performed using a cause-specific hazards model. Differences between intervention groups will be tested using Gray's log-rank test.
Potential heterogeneity of the treatment effect will be assessed based on appropriate interaction (likelihood ratio) tests and the following pre-defined sub-grouping variables in separate models: • Pre-existing medical conditions

• Duration of prone position
For other secondary outcome, all analyses will be performed for the comparison of intervention vs. standard care. For continuous outcomes such as vital signs (SpO 2 /respiratory rate/ heart rate/ FiO 2 ), supplemental oxygen free days, etc., interventions will be compared using linear regression, but we will add a quadratic term if it gives the model a better fit. For dichotomous outcomes such as requirement for intubation and mechanical ventilation within 28 days of randomization, we will compute the number of patients who developed or did not develop the outcomes of interest and we fit a logistic regression. Time-to-event analysis will be performed using a cause-specific hazards model. Differences between intervention groups will be tested using Gray's log-rank test. For all time-to-event analysis patients that withdrew will be censored at the time of withdrawal. Mortality will be visualised in each arm using Kaplan-Meier curves and modelled using Cox regression.

Data management
Source data Source documents are where data are first recorded, and from which participants' CRF data are obtained. These include, but are not limited to, hospital records (from which medical history and previous and concurrent medication may be summarised into the CRF), clinical and office notes (including electronic health record), laboratory and pharmacy records, radiographs, and correspondence. CRF entries will be considered source data if the CRF is the site of the original recording (e.g. there is no other written or electronic record of data). All documents will be stored safely in confidential conditions. On all trial-specific documents, other than the signed consent, the participant will be referred to by the trial participant number/code, not by name.

Direct access to participant records
Participating investigators should agree to allow trial-related monitoring, including audits, ethics committee review and regulatory inspections by providing direct access to source data and documents as required. Participants' consent for this must be obtained. Such information will be treated as strictly confidential and will in no circumstances be made publicly available.
The following data should be verifiable from source documents: • all signed consent forms • dates of assessments including dates specimens were taken and processed in the laboratory Data will be entered a 21 CFR Part 11-compliant data capture system. The data system includes password protection and internal quality checks, such as automatic range checks, to identify data that appear inconsistent, incomplete, or inaccurate. Clinical data will be entered directly from the source documents.
The participants will be identified by a unique trial specific number and/or code in any database. The name and any other identifying detail will NOT be included in any trial data electronic file.

Trial records retention
CRFs, clinical notes and administrative documentation will be kept in a secure location and held for 15 years after the end of the trial. Clinical information will not be released without written permission, except as necessary for monitoring, auditing and inspection purposes. During this period, all data should be accessible to the competent authorities with suitable notice. Electronic data will be kept for at least 20 years at the OUCRU CTU.

Protocol violations
A protocol deviation is any non-compliance with the clinical trial protocol or GCP requirements. If such a deviation results in an impact on patient safety or scientific integrity it becomes a protocol violation. The non-compliance may be either on the part of the participant, the investigator, or the study site staff. Whenever violations occur, corrective actions are to be developed by the site and implemented promptly.
It is the responsibility of the site investigators to use continuous vigilance to identify and report protocol deviations and violations. All deviations and violations must be documented in source documents and reported to the OUCRU CTU within 2 days of being identified. In addition, protocol violations must be reported to the relevant ethics committees.

Publication and data sharing policy
All publications are to be approved by the trial steering committee before submission for publication. Any publication arising before all patients have completed follow-up (not by randomized groups) will also be approved by the DSMB in order to ensure that the primary objective of the trial (the randomized comparison) is not compromised.
In line with Wellcome Trust policy that the results of publicly-funded research should be freely available, manuscripts arising from the trial will be published in a Plan-S compliant manner. All publications will acknowledge the trial's funding sources.
In line with research transparency and greater access to data from trials OUCRU's clinical trials are registered at Clinical-Trials.gov and a data sharing policy is in place. This policy is based on a controlled access approach with a restriction on data release that would compromise an ongoing trial or study.
Data exchange complies with Information Governance and Data Security Policies in all of the relevant countries.

Risk assessment
The Quality Assurance (QA) and Quality Control (QC) considerations have been based on a formal Risk Assessment, which acknowledges the risks associated with the conduct of the trial and how to address them with QA and QC processes. QA includes all the planned and systematic actions established to ensure the trial is performed and data generated, documented and/or recorded and reported in compliance with the principles of ICH GCP and applicable regulatory requirements. QC includes the operational techniques and activities done within the QA system to verify that the requirements for quality of the trial-related activities are fulfilled.
The safety profile of prone position is well-known and acceptable, given the potential benefits. The trial will be recruiting sick participants, but site investigators have considerable experience with this population. This will minimise the risks to the participants and the trial. A detailed risk assessment will be conducted prior to starting the trial.
Central monitoring at OUCRU CTU A site initiation visit will be conducted for each study site by staff from the OUCRU CTU. All essential site staff must be in attendance. On site monitoring will also be regularly conducted by the site monitors. The frequency, type and intensity of routine monitoring and the requirements for triggered monitoring will be detailed in the Monitoring Plan which will also detail the procedures for review and sign-off. The monitoring will adhere to the principles of ICH GCP and the Monitoring Plan.
The monitors will require access to all participant medical records including, but not limited to, laboratory test results and prescriptions. The investigator (or delegated deputy) should work with the monitor to ensure that any problems detected are resolved.

Regulatory and ethical considerations
All regulatory requirements will be met by the co-sponsors or their delegated authorities.

Compliance
The trial end is 28 days after the last participant is randomized (end of follow-up for the last randomized participant).
The trial complies with the principles of the Declaration of Helsinki (2008) and will be conducted in compliance with the approved protocol and the principles of Good Clinical Practice (GCP).
All sites will comply with the above. An agreement will be in place between the site and the OUCRU CTU, setting out respective roles and responsibilities.
The site will inform the CTU as soon as they are aware of a possible serious breach of compliance. For the purposes of this regulation, a 'serious breach' is one that is likely to affect to a significant degree the safety or physical or mental integrity of the subjects in the trial, or the scientific value of the trial Ethical conduct of the study Ethical considerations All participants will receive the best available treatment of COVID-19, following local and national guidelines. They will benefit from the frequent and careful follow-up of their condition throughout the treatment of their disease and for up to 28 days from randomization.
The risks and benefits of participation will be communicated in two ways. First, all potential participants or their family members will be given a participant information sheet clearly listing the risks and benefits of the trial. Second, all potential participants (or their families) will be able to discuss participation with their consulting doctor who will be able to address questions not covered or arising from the participant information sheet.
The trial protocol will seek ethical approval to include incapacitated, comatose adults in the trial as we consider many of these adults will have the most severe disease and therefore represents the group that might stand most to gain from the intervention.
Participants' confidentiality will be maintained throughout the trial. Data submitted to OUCRU CTU and samples sent to central testing facilities will be identified only by the trial number and participant initials.

Ethical approvals
The trial will be approved by the Oxford Tropical Research and local Ethics Committee.
Any further amendments will be submitted and approved by the relevant ethics committee.
The rights of the participant to refuse to participate in the trial without giving a reason must be respected. After the participant has entered into the trial, the clinician must remain free to give alternative treatment to that specified in the protocol, at any stage, if he/she feels it to be in the best interest of the participant. The reason for doing so, however, should be recorded; the participant will remain within the trial for the purpose of follow-up and for data analysis by the treatment option to which they have been allocated. Similarly, the participant must remain free to change their mind at any time about the protocol treatment and trial follow-up without giving a reason and without prejudicing his/her further treatment.

Confidentiality
The investigator must assure that participants' anonymity will be maintained and that their identities are protected from unauthorised parties. Participants will be assigned a trial identification number and this will be used on CRFs; participants will not be identified by their name. The investigator will keep securely a participant trial register showing identification numbers, surnames and date of birth. This unique trial number will identify all laboratory specimens, case record forms, and other records and no names will be used, in order to maintain confidentiality.

Expenses
Treatment and hospital costs from enrolment to discharge from hospital for all actively enrolled participants will be covered by the State Budget.
The study will not cover the cost of treating pre-existing diseases or those unrelated to study participation or the diagnosis and/or treatment of COVID-19.

Oversight and trial committees
Trial management group (TMG) A Trial Management Group (TMG) will be formed to conduct the day-to-day management of the trial at the OUCRU CTU. This will include the Chief Investigator, Trial Statistician, Clinical Project Manager, Trial Manager and Data Manager. The group will meet at least once per month, although may meet more or less often as required. The group will discuss issues related to the progress of the trial at the site and to ensure that the trial is running well. The full details can be found in the TMG Charter.

Trial steering committee (TSC)
The Trial Steering Committee (TSC) has membership from the TMG plus other members. The role of the TSC is to provide overall supervision for the trial and provide advice. The ultimate decision for the continuation of the trial lies with the TSC. Further details of TSC functioning are presented in the TSC Charter.

Data management and safety board (DSMB)
An independent DSMB will oversee the safety of the trial. The DSMB will be the only group which sees the confidential, accumulating data for the trial separately by randomized group. A DSMB Charter will be drawn up that describes the membership of the DSMB, relationships with other committees, terms of reference, decision-making processes, and the approximate timing and frequency of interim analyses (with a description of stopping rules and/or guidelines). In addition, an interim analysis plan will be written which details all statistical tables that will be provided to the DSMB.

Insurance
The conduct of this study is sponsored by the University of Oxford. The University has a specialist insurance policy in place: -Newline Underwriting Management Ltd, at Lloyd's of London -which would operate in the event of any participant suffering harm as a result of their involvement in the research.

Discussion
Whilst significant progress has been made towards understanding the role of APP in moderate to severe COVID-19, the evolving pandemic means that our study can make unique contributions to further our understanding of the value of this intervention, particularly in limited-resource settings.
Although the majority of cases of COVID-19 are currently in LMICs 17 , almost all the evidence for interventions comes from higher income settings. Evidence of APP efficacy in LMICs principally comes from Ehrmann et al's study which included 430 patients from Mexico recruited before January 2021 4 . Our study, however, captures data from a largely vaccinated population in a country with a significantly lower per-capita healthcare expenditure (180 USD compared to 540 USD in Mexico 18 ). Observation from initial recruitment is that our study will enrol many patients with immunosuppression, including HIV. Previous studies carried out early in the pandemic mainly describe unvaccinated populations and comorbidities related to cardiovascular and pulmonary disease. Understanding the role of APP in the current context however is particularly important for LMICs where HIV remains common, as do other practices such as long-term unrestricted steroid use in the community.
At present, daily cases of COVID-19 in Ho Chi Minh City are lower than when our study was conceived, and consequently we expect a slower recruitment rate 19 . This offers several advantages. Firstly, with improved availability of monitoring equipment and staff: patient ratios, we believe that we will be able to collect higher quality routine data with less missing data. Secondly, even though we have a specific team to deliver the intervention, they will be able to do this more carefully and thoroughly ensuring high compliance and accurate recording of any adverse events.
We believe our study is important in that we are employing a special team to deliver the proning intervention. This addresses the issue of low compliance noted in previous studies and postulated as a reason for failure of APP in many 9,20,21 . Assessing compliance has been further complicated by considerable heterogeneity of methods used to record prone position duration in preceding studies. Some studies have used selfreported methods, whilst some others have not described methods clearly (or at all) 20,22,23 . Our study design enables us to accurately quantify the duration of prone position using a combination of regular study-team evaluated observations and continuous monitoring with wearable gyrometry devices.
The use of wearable devices in this study is also a new innovation for resource-limited settings. In addition to gyrometry, our use of wearable vital sign monitoring devices alleviates problems associated with equipment availability, enabling continuous monitoring and as well as remote monitoring for either routine clinical purposes or study-related observations. The formal finalized statistical analysis plan will be published before trial analysis, however we will also be able to carry out exploratory analysis using data from wearable devices. Whilst previously tested in routine ICU care 24 , this study will provide proof-of-principal data on utility for high-consequence infectious diseases in limited-resource settings.

Ethics approvals
The trial has been approved by the Scientific and Ethics Committee of the Hospital for Tropical Diseases, Ho Chi Minh City (4178/QD-BVBND) and the Oxford Tropical Research Ethics Committee (OxTrec) (39-21)

Clinical Trial Registration
The trial is registered at Clinical Trials.GoV NCT05083130

Data (and software) availability
These will be available in a timely manner according to Wellcome Trust policy (see protocol)

Reporting guidelines
The trial will be reported according to CONSORT guidelines.  Participant timeline 13 Time schedule of enrolment, interventions (including any run-ins and washouts), assessments, and visits for participants. A schematic diagram is highly recommended (see Figure) 10,11 Methods for any additional analyses (eg, subgroup and adjusted analyses) 14 20c Definition of analysis population relating to protocol non-adherence (eg, as randomised analysis), and any statistical methods to handle missing data (eg, multiple imputation) 14

Methods: Monitoring
Data monitoring 21a Composition of data monitoring committee (DMC); summary of its role and reporting structure; statement of whether it is independent from the sponsor and competing interests; and reference to where further details about its charter can be found, if not in the protocol. Alternatively, an explanation of why a DMC is not needed 19 21b Description of any interim analyses and stopping guidelines, including who will have access to these interim results and make the final decision to terminate the trial How personal information about potential and enrolled participants will be collected, shared, and maintained in order to protect confidentiality before, during, and after the trial 18 Declaration of interests 28 Financial and other competing interests for principal investigators for the overall trial and each study site 19 Access to data 29 Statement of who will have access to the final trial dataset, and disclosure of contractual agreements that limit such access for investigators Section/item

ItemNo Description Page
Ancillary and post-trial care 30 Provisions, if any, for ancillary and post-trial care, and for compensation to those who suffer harm from trial participation 18 Dissemination policy 31a Plans for investigators and sponsor to communicate trial results to participants, healthcare professionals, the public, and other relevant groups (eg, via publication, reporting in results databases, or other data sharing arrangements), including any publication restrictions