You are here

  • Home
  • Archive
  • Volume 11, Issue 1
  • Automated oxygen titration with non-invasive ventilation in hypoxaemic adults with cardiorespiratory disease: a randomised cross-over trial

Article Text

Article menu

Download PDFPDF

Non-invasive ventilation
Automated oxygen titration with non-invasive ventilation in hypoxaemic adults with cardiorespiratory disease: a randomised cross-over trial
  1. Louis Kirton1,2,
  2. Stacey Kung1,
  3. Georgina Bird1,
  4. Melissa Black1,
  5. Ruth Semprini1,
  6. Allie Eathorne1,
  7. Mark Weatherall3,
  8. Alex Semprini1,2 and
  9. Richard Beasley1,2
  1. 1Medical Research Institute of New Zealand, Wellington, New Zealand
  2. 2Victoria University Wellington, Wellington, New Zealand
  3. 3University of Otago Wellington, Wellington, New Zealand
  1. Correspondence to Dr Louis Kirton; louis.kirton{at}mrinz.ac.nz

Abstract

Background Closed-loop oxygen control systems automatically adjust the fraction of inspired oxygen (FiO2) to maintain oxygen saturation (SpO2) within a predetermined target range. Their performance with low and high-flow oxygen therapies, but not with non-invasive ventilation, has been established. We compared the effect of automated oxygen on achieving and maintaining a target SpO2 range with nasal high flow (NHF), bilevel positive airway pressure (bilevel) and continuous positive airway pressure (CPAP), in stable hypoxaemic patients with chronic cardiorespiratory disease.

Methods In this open-label, three-way cross-over trial, participants with resting hypoxaemia (n=12) received each of NHF, bilevel and CPAP treatments, in random order, with automated oxygen titrated for 10 min, followed by 36 min of standardised manual oxygen adjustments. The primary outcome was the time taken to reach target SpO2 range (92%–96%). Secondary outcomes included time spent within target range and physiological responses to automated and manual oxygen adjustments.

Results Two participants were randomised to each of six possible treatment orders. During automated oxygen control (n=12), the mean (±SD) time to reach target range was 114.8 (±87.9), 56.6 (±47.7) and 67.3 (±61) seconds for NHF, bilevel and CPAP, respectively, mean difference 58.3 (95% CI 25.0 to 91.5; p=0.002) and 47.5 (95% CI 14.3 to 80.7; p=0.007) seconds for bilevel and CPAP versus NHF, respectively. Proportions of time spent within target range were 68.5% (±16.3), 65.6% (±28.7) and 74.7% (±22.6) for NHF, bilevel and CPAP, respectively.

Manually increasing, then decreasing, the FiO2 resulted in similar increases and then decreases in SpO2 and transcutaneous carbon dioxide (PtCO2) with NHF, bilevel and CPAP.

Conclusion The target SpO2 range was achieved more quickly when automated oxygen control was initiated with bilevel and CPAP compared with NHF while time spent within the range across the three therapies was similar. Manually changing the FiO2 had similar effects on SpO2 and PtCO2 across each of the three therapies.

Trial registration number ACTRN12622000433707.

  • Non invasive ventilation

Data availability statement

Data are available on reasonable request. Anonymised datasets are available on reasonable request, until a minimum of 10 years after publication to researchers who provide a methodologically sound proposal that has been approved by the study investigators. This is possible through a signed data access agreement and subject to approval by the principal investigator LK (louis.kirton@mrinz.ac.nz) and the study sponsor James Revie (James.Revie@fphcare.co.nz)

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjresp-2023-002196

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Automated oxygen control systems that titrate oxygen to a target oxygen saturation (SpO2) range have been developed for use with conventional low-flow and nasal high-flow (NHF) oxygen therapies. Their use with bilevel and continuous positive airway pressure (CPAP) therapies in patients with cardiorespiratory disease is yet to be studied.

    WHAT THIS STUDY ADDS

    • In hypoxaemic participants with stable cardiorespiratory disease, applying automated oxygen control with bilevel and CPAP therapies took less time to reach the target SpO2 range than with NHF therapy while the time spent within the target range was similar across each therapy. When manually adjusting the inspired oxygen concentration (FiO2) using NHF, bilevel and CPAP; standardised increments and decrements in the FiO2 resulted in similar rises and falls in SpO2 and transcutaneous carbon dioxide levels, respectively, across each of the three therapies.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • These findings suggest that automated oxygen control systems with bilevel and CPAP are at least as effective as NHF in achieving oxygen titration within a target SpO2 range when used in participants with cardiopulmonary disease and are suitable for use in those with acute respiratory failure who require these therapies.

    Introduction

    In acutely unwell patients with hypoxia, international guidelines recommend that oxygen is titrated to a predetermined peripheral oxygen saturation (SpO2) target.1–3 This targeted approach to oxygen delivery is potentially advantageous to patients as it mitigates the known harms of both hypoxaemia and hyperoxaemia.1 4 5 The current process for delivering oxygen to an SpO2 target range is predominantly manual and is performed with variable accuracy between clinicians and across hospital settings. In a standard medical ward, adjusting oxygen manually appears suboptimal, with SpO2 target range maintained only around half the time.6 This compares to intensive care units (ICUs), where manually adjusted oxygen under clinical trial conditions has been reported to maintain SpO2 within target range approximately 90% of the time,7 while in a cohort study, adherence to the target SpO2 range was achieved only 43% of the time.8

    Novel automated oxygen control systems have been developed, operating an algorithm-based closed-loop controller that automatically titrates the delivered fractional concentration of inspired oxygen (FiO2) to maintain SpO2 within a prespecified target range. The effect of these automated oxygen control systems using both low-flow and high-flow oxygen devices is to markedly improve the efficacy of targeted oxygen delivery.7 9–13 When trialled in the emergency department and following abdominal or thoracic surgery, automated low-flow oxygen systems maintain target range saturation more effectively than manual oxygen control.10 11 Similarly, when used with nasal high flow (NHF) in a standard medical ward and in an ICU,7 9 automated oxygen control improved adherence to a prescribed saturation target range.

    An automated oxygen control system integrated into a non-invasive ventilation (NIV) device represents a novel application of this technology,14 that is, supported by only one randomised trial in healthy volunteers.15 Further evidence is needed to characterise how automating oxygen delivery affects adherence to an SpO2 target range while using NIV therapy in participants with underlying cardiopulmonary disease.

    Accordingly, this trial examines a novel device (Airvo-3 NIV with OptiO2, Fisher and Paykel Healthcare, Auckland, New Zealand) capable of automating oxygen delivery across each of its three respiratory support platforms: NHF, bilevel and continuous positive airway pressure (CPAP). As each treatment modality influences the respiratory system in a distinct way, generating different conditions for ventilation and gas exchange,16 it is necessary to assess and compare the efficacy of automated oxygen control using each mode. The main objective of this study was to estimate the effect of initiating NHF, bilevel and CPAP therapies while using automated oxygen control on time taken to restore target range saturations in adults with resting hypoxaemia. The hypothesis was that the time taken for SpO2 to come into range would be shorter when used with bilevel and CPAP as compared with NHF. Additionally, we aimed to characterise the effect of manual changes in the FiO2 with NHF, bilevel and CPAP on SpO2 and transcutaneous carbon dioxide (PtCO2).

    Methods

    Study design

    This was an open-label, three-way cross-over, randomised clinical trial undertaken at the Medical Research Institute of New Zealand facility in Wellington Hospital. A cross-over design was chosen for this study because the within-patient variation is less than the between-patient variation meaning fewer participants were required. The three interventions (NHF, bilevel and CPAP) lasted 46 min each. Each 46 min intervention consisted of two parts; first a 10 min period of automatically adjusted oxygen where the target SpO2 range was set to 92%–96%, and second a 36 min period where the FiO2 was adjusted manually every 6 min in a standardised sequence as follows: 0.21, 0.25, 0.32, 0.45, 0.25 and 0.21. A washout period of at least 20 min occurred between each intervention which has been shown to provide sufficient time for SpO2 and PtCO2 to return to baseline17 (figure 1).

    Figure 1
    Figure 1

    Schematic representation of study procedures. aContinuous measurements include: Transcutaneous carbon dioxide (PtCO2; SenTec), peripheral saturation of oxygen (SpO2; Nonin 7000A and SenTec), heart rate (Nonin) and respiratory rate (Airvo-3 NIV). Bilevel, bilevel positive airway pressure; CPAP, continuous positive airway pressure; FiO2, fraction of inspired oxygen; NHF, nasal high flow.

    This trial was registered with the Australian and New Zealand Clinical Trials Registry, ACTRN12622000433707.

    Patient and public involvement

    The study design was reviewed and endorsed by the Research Advisory Group—Māori (Māori Partnership Board, Capital and Coast District Health Board, reference number RAG-M #927).

    Participants

    Participants were adults with stable respiratory or cardiovascular disease and resting SpO2≤90% while breathing room air. Exclusion criteria were a contraindication to NIV as per British Thoracic Society/Intensive Care Society guidelines,18 known infection or colonisation with multidrug-resistant bacteria, SARS-CoV-2 infection, pregnancy or any other condition believed to be a safety risk (at investigators discretion). All participants gave written informed consent.

    Randomisation

    Participants were randomised in equal proportions to one of six possible orders to receive the three interventions (a-b-c, a-c-b, b-a-c, b-c-a, c-a-b and c-b-a). The randomisation code was generated by the study statistician using a computer-generated sequence. Allocation was concealed by the SMART-TRIAL electronic case report form and was released to investigators at the time of randomisation. The nature of the intervention meant participants could not be blinded to the allocated order of the interventions.

    Procedures

    Participants attended a single visit. Resting SpO2 was recorded using a disposable finger probe pulse oximeter (Nonin 7000A, Minnesota, USA), with a 10 min rolling mean value used to assess eligibility. Forced expiratory volume in 1 s and functional vital capacity were measured prior to the trial interventions (Easy on-PC, NDD Medical Technologies, Zurich, Switzerland). Spirometry was performed according to American Thoracic Society/European Respiratory Society criteria.19

    Measurements of PtCO2, SpO2 and heart rate (HR) were captured using an ear probe sensor from a SenTec (SenTec AG, Therwil (Basel), Switzerland) digital monitor. Finger, cheek and forehead pulse oximeters (Nonin) were also fitted to continuously monitor SpO2. Blood pressure (BP) was monitored intermittently (Mindray ePM 12, Shenzhen, China).

    All participants then proceeded through the three interventions in randomly allocated order as outlined in figure 1. NHF was delivered using an AirSpiral heated breathing tube (Fisher & Paykel Healthcare) and Optiflow+nasal cannula (Fisher & Paykel Healthcare) at 35 L/min flow, with an initial temperature of 37°C. Bilevel was applied in spontaneous-timed mode using a Nivairo (Fisher & Paykel Healthcare) full face mask, with inspiratory positive airway pressure 15 cmH20, expiratory positive airway pressure 5 cmH20, minimum inspiratory time 1.0 s, backup respiratory rate (RR) 12 and an initial temperature of 27°C. CPAP was applied using a Nivairo full face mask, with 10 cmH20 CPAP and an initial temperature of 27°C.

    During the first 10 min automated oxygen control part of each intervention, the SpO2 target range was set at 92%–96%. It was deemed by the trial management committee that even though a saturation target range of 88%–92% may be more physiologically appropriate for the study population, the effect of the automated oxygen control system would be best observed with a range of 92%–96% and potential transient exposure to SpO2 levels above the upper limit while under strict observation was considered acceptable.

    Outcomes

    The study’s primary outcome was the mean (±SD) time taken for the SpO2 to reach target range, of between 92% and 96% while using the automated oxygen control system. This was defined as the time (in seconds) between activating the Airvo-3 device to deliver NHF, bilevel or CPAP therapy, and the first moment when the participants SpO2 was ≥92%.

    Secondary outcomes included the time taken to reach target range compared across the three modes of respiratory support, the proportion of time spent within target range and the time course of SpO2, FiO2 and transcutaneous carbon dioxide (PtCO2) during the 10 min period of the automated oxygen control using each mode of therapy. Additional secondary outcomes included maximum, minimum and median SpO2, HR, RR, BP and PtCO2 over each 6 min interval during the manual FiO2 ramp phase, as well as the mean difference in SpO2 between the finger probe pulse oximeter and ear-lobe oximeters (together with limits of agreement) at the beginning of the NHF intervention.

    Sample size

    At the time of study design, an estimate of paired SD for time-to-target-range was not available and so the study sample size was based on having reasonable precision to estimate mean time-to-target range. This was based on a known mean and SD for single treatment time-target range of 120 and 24 s, respectively. A sample size of 12 allows for a 95% CI for a single modality of treatment time-to-target range of plus or minus 20 s.

    Statistical analysis

    For the primary outcome variable, a linear mixed model was used to compare the three treatments, with participant treated as a random effect to account for the same participants being measured three times, and order of treatment and treatment modality as fixed effects. Similar linear mixed models were used to compare percentage of time in, below and above target SpO2 range. PtCO2 was also compared between treatments using a linear model, with baseline PtCO2 as an additional fixed effect. The time course of physiological variables is shown graphically as scatter plots of raw and paired data with locally estimated scatter plot smoother and approximate 95% CI.

    For the secondary outcome variables of HR, systolic BP (SBP), diastolic BP (DBP) and RR, a linear mixed model was used to compare the three treatments, with participant treated as a random effect to account for the same participants being measured three times, and order of treatment and treatment modality as fixed effects.

    Agreement between pulse oximeters was compared using the NHF data, by linear mixed model with fixed effects for the type of pulse oximeter and random effects for both participant and algorithm step. The mean difference between pulse oximeters was tested for evidence of bias and limits of agreement were estimated by two times the square root of the residual variance in the linear mixed model. Physiological variables across the 6 min intervals of the FiO2 manual ramp phase are presented as mean and standard deviation (±SD) and median and IQR and minimum to maximum.

    SAS V.9.4 (SAS Institute) was used.

    Results

    Participants were recruited between 7 April 2022 and 27 July 2022. 14 participants were screened for the trial with 2 subjects excluded because of a baseline SpO2>90% (figure 2). The 12 participants who completed the trial comprised of 7 females and 5 males, with the prevalence of active cardiorespiratory diagnoses distributed as follows; chronic obstructive pulmonary disease (10/12), heart failure (7/12), obstructive sleep apnoea (OSA) (3/12), OSA/obesity hypoventilation syndrome (2/12), diffuse parenchymal lung disease (2/12), bronchiectasis (1/12). 10 participants were prescribed long-term oxygen therapy, and three participants were prescribed nocturnal positive airway pressure therapy (bilevel or CPAP) (table 1). Two participants were randomised to each of the six potential treatment orders.

    View this table:
    Table 1

    Participant characteristics

    Figure 2
    Figure 2

    CONSORT flow diagram. (A) Bilevel, bilevel positive airway pressure; (B) CPAP, continuous positive airway pressure; (C) NHF, nasal high flow; PtCO2, transcutaneous carbon dioxide; SpO2, peripheral saturation of oxygen; SBP, systolic blood pressure.

    Automated oxygen control period

    The mean (±SD) time to reach target SpO2 was 114.8 (±87.9), 56.6 (±47.7) and 67.3 (±61) seconds for NHF, bilevel and CPAP, respectively, with a mean difference (95% CI) of 58.3 s (25.0 to 91.5), p=0.002; and 47.5 s (14.3 to 80.7), p=0.007; for bilevel and CPAP versus NHF, respectively (table 2). The trajectories of the SpO2 plots for bilevel and CPAP immediately separated from that of NHF and tracked almost identically across the 10 min period of automated oxygen control (figure 3A). As the SpO2 plots tracked into target range, the corresponding FiO2 plot shows different FiO2 trajectories by therapy, with a gradual increase while using NHF and a gradual decrease while using bilevel and CPAP (figure 3B).

    View this table:
    Table 2

    Automated oxygen control period—time taken to reach SpO2 target range, and time spent below, above and within range

    Figure 3
    Figure 3

    LOESS plots of: (A) SpO2, (B) FiO2 and (C) PtCO2, over time by treatment over the 10 min automated oxygen control period. Locally estimated scatterplot smoothing (LOESS), where the darker bold line in the centre of the coloured band represents the predicted SpO2, FiO2 or PtCO2 and the upper and lower limits of the coloured band represent the surrounding 95% CI. Bilevel, bilevel positive airway pressure; CPAP, continuous positive airway pressure; FiO2, fraction of inspired oxygen; NHF, nasal high flow; PtCO2, transcutaneous carbon dioxide; SpO2, peripheral oxygen saturation.

    Secondary outcomes relevant to the 10 min automated oxygen control period include reduced time spent below SpO2 target range for both bilevel and CPAP when compared with NHF, increased time spent above range for bilevel compared with NHF, no significant difference in time spent above range comparing CPAP and NHF, and no significant differences in time spent within range across all three treatments (table 2). The magnitude of increase in SpO2 during the automated oxygen control period varied between individuals; with some overshooting the 96% upper limit of the SpO2 target range. Line plots of SpO2 and FiO2 by individual show that at times where the SpO2 increased above 96%, the FiO2 delivered by the closed-loop oxygen controller was at or near 0.21 (see online supplemental figure S1–S9).

    PtCO2 trended downward over the 10 min automated oxygen period as participants were initiated on bilevel while remaining near constant in participants receiving NHF or CPAP (online supplemental table S1, figure 3C).

    Manual oxygen adjustment periods

    Over each 6 min step during all three treatments, manual sequential FiO2 increments and decrements were associated with a rise and fall in SpO2, respectively. While the mean change in SpO2 associated with an increment or decrement in FiO2 was equivalent across each treatment, the baseline SpO2 for the pressure-based therapies (bilevel and CPAP) was higher than for NHF, a separation that was maintained at each stage of the manual oxygen control period (figure 4A).

    Figure 4
    Figure 4

    LOESS plots of: (A) FiO2 and (B) PtCO2 over time by treatment over the manual oxygen control period. Locally estimated scatterplot smoothing (LOESS), where the darker bold line in the centre of the coloured band represents the predicted FiO2 or PtCO2 and the upper and lower limits of the coloured band represent the surrounding 95% CI. Bilevel, bilevel positive airway pressure; CPAP, continuous positive airway pressure; FiO2, fraction of inspired oxygen; NHF, nasal high flow; PtCO2, transcutaneous carbon dioxide.

    Mean PtCO2 increased with each FiO2 increment (steps 2–5) and decreased with each FiO2 decrement (steps 6–7), with aberrances in this trend evident during CPAP at step 3 where one participant was withdrawn for consecutive low SBPs (<90 mm Hg), during bilevel at step 5 where one participant’s PtCO2 increased from 56 mm Hg to 67 mm Hg (figure 4B).

    Other variables measured included HR, SBP, DBP and RR. During the manual FiO2 adjustment periods, the SBP and DBP were significantly lower comparing bilevel to NHF, but with no difference between CPAP and NHF, when differences were averaged across all steps (see online supplemental table S2 and S3). One participant was withdrawn part way through each of the three treatments for consecutive SBP<90 mm Hg. During the manual FiO2 adjustment periods, HR was lower comparing bilevel to NHF when averaged across all stages (see online supplemental table S4). Inaccuracies in RR sensing during the NHF treatment period prohibited meaningful comparison across treatments.

    Bias was observed between the finger probe Nonin (7000A, Minnesota, USA) pulse oximeter and the earlobe SenTec (SenTec AG, Therwil (Basel), Switzerland) sensor, with significantly higher SpO2 values recorded by the latter. The mean (95% CI) difference was −1.90% (−2.28% to −1.51%), p<0.0001 and limits of agreement −4.17% to 0.37% (see online supplemental figure S10).

    Discussion

    In participants with advanced cardiorespiratory disease and resting hypoxaemia, SpO2 increased to within the target SpO2 range more quickly when initiating the OptiO2 automated oxygen control system with bilevel and CPAP than with NHF. The more rapid restoration of target saturation range meant that the proportion of time when SpO2 was below target range was less for CPAP and bilevel, than NHF. These findings align with the known effects of initiating bilevel and CPAP therapies, whereby the positive airway pressure promptly increases above what can be achieved with NHF,17 20 resulting in a more rapid improvement in gas exchange. Over the 10 min automated oxygen control period, time spent within the target SpO2 range using each mode of therapy was not different, suggesting that automated oxygen control systems with bilevel and CPAP are at least as effective as with NHF.

    The average increase in SpO2 from baseline was greater when initiating the OptiO2 system with bilevel and CPAP, compared with NHF. In some individuals, the rise in SpO2 was more pronounced, breaching the 96% upper limit of the SpO2 target range. When tracking the time course of SpO2 and FiO2 by individual participant over the automated oxygen control period, it appears that on each occasion when SpO2 exceeded 96%, the delivered FiO2 was at, or close to 0.21, meaning that this level of oxygenation was likely achieved because of improved pulmonary mechanics and gas exchange imparted by pressure or flow therapies,16 17 20–24 rather than an automated oxygen controller unable to enact a timely reduction of delivered FiO2. While differences in the rate and degree of increase in SpO2 were observed when the different therapies were initiated, the overall proportion of the 10 min automated oxygen control period when SpO2 was inside the target range was similar between each therapy. This implies that despite differences in SpO2 responsiveness by therapy, the automated oxygen controller adapted appropriately so that target range saturations were maintained with near equivalence.

    Trends in PtCO2 identified changes in ventilation status in response to trial therapies. Over the 10 min automated oxygen control period with bilevel therapy the mean PtCO2 decreased signifying an increase in alveolar ventilation that was not observed while using CPAP and NHF. This finding is compatible with the well-established ventilatory effects of each therapy, with bilevel providing pressure support to directly increase alveolar ventilation25 and with the positive end-expiratory pressure generated by CPAP and NHF having only a minor and indirect impact on ventilation.16 For all three therapies, as participants progressed into the manual FiO2 adjustment periods, PtCO2 trended upward as the FiO2 was stepped up to 0.45. This may be a consequence of hyperoxaemia-induced hypercapnia,21 26 27 a phenomena that arises from a transient reduction in ventilatory drive, with a sustained rise in arterial carbon dioxide,28 as well as reversal of hypoxic pulmonary vasoconstriction resulting in increased dead space.29 30 Notably, excessive oxygenation in critically unwell patients using bilevel NIV has previously been reported to have no significant effect on arterial carbon dioxide, minute ventilation and dead space31 so our contrary finding that PtCO2 trended upward when participants received escalating oxygen concentrations with bilevel NIV requires further exploration.

    To our knowledge, this is the first trial of its kind whereby an automated oxygen control system was tested across multiple respiratory support treatments in chronic cardiopulmonary disease, broadening the clinical situations in which this technology has been successfully applied. Moreover, this trial examined the efficacy of the automated oxygen control system by restoration of normoxaemia in resting patients with baseline hypoxaemia, rather than the more conventional approach of assessing how the automated oxygen controller responds to exercise-induced hypoxaemia.15 32–34 This allowed for simultaneous evaluation of the FiO2 controller’s response to hypoxaemia under therapy-specific pressure and flow conditions, as well as to high SpO2 values that were driven above the 96% upper limit of the target range by application of respiratory support.

    While this trial captures favourable performance characteristics of the automated oxygen control system, the static nature of the testing and short 10 min testing period of the automated oxygen control function limits the applicability of the findings, as real-life situations where the Airvo-3 NIV device with OptiO2 might best be used may pose more pronounced and enduring periods of hypoxaemia or hyperoxaemia.6 While participants received automated oxygen control, a single SpO2 range of 92%–96% was targeted so our findings may not be generalisable to other target ranges. Additionally, when participants received manual oxygen control, stepwise changes in FiO2 were made at 6 min intervals, which does not reflect real-world clinician-initiated oxygen adjustment. Further studies in acutely unwell hypoxaemic patients, over longer durations, with different target SpO2 ranges are required to understand the effects of automated oxygen control systems under acute care conditions.

    A recent trial reports that automated oxygen control using low-flow oxygen is likely to improve endurance capacity and dyspnoea scores compared with manually adjusted oxygen, with no additional benefit from the use of automated oxygen control with NHF.35 A limitation of our trial was that participants were at rest throughout the entirety of the intervention so an assessment of exertional breathlessness and endurance capacity while using NHF, bilevel and CPAP, in automated oxygen control mode was not appropriate. Additionally, our trial did not directly compare automated and manual oxygen control on the time taken to reach, or the time spent within, the target SpO2 range.

    Measurement bias was observed between the Nonin finger and Sentec ear-lobe pulse oximeters. In this study, the limits of agreement with gold-standard arterial SpO2 (%SaO2) could not be assessed, as no arterial blood gas sampling was performed, but given the relatively wide limits of agreement measured (−4.17 to 0.37), errors in SaO2 approximation according to oximeter site and brand are likely to significantly influence accuracy of the automated oxygen control system. Further exploration of patient variables that influence pulse oximetry, as well as characterisation of brand-specific oximeter performance are required so that automated oxygen control systems can more accurately maintain target SpO2.

    While there were distinct differences in the performance of the OptiO2 automated oxygen control system across the NHF, bilevel and CPAP interventions, the findings of this trial indicate that this system is highly effective at maintaining SpO2 within target range across each of the tested treatment platforms. Based on these findings, clinical trials of its use in acute care in patients with respiratory failure requiring bilevel and/or CPAP are indicated.

    Data availability statement

    Data are available on reasonable request. Anonymised datasets are available on reasonable request, until a minimum of 10 years after publication to researchers who provide a methodologically sound proposal that has been approved by the study investigators. This is possible through a signed data access agreement and subject to approval by the principal investigator LK (louis.kirton@mrinz.ac.nz) and the study sponsor James Revie (James.Revie@fphcare.co.nz)

    Ethics statements

    Patient consent for publication

    Ethics approval

    Ethical approval was obtained from the Northern B Health and Disability Ethics Committee (Review reference 2021 FULL 11508). The trial was run in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki.

    References

    1. ↵
      1. O’Driscoll BR,
      2. Howard LS,
      3. Earis J, et al
      . BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax 2017;72:ii1–90. doi:10.1136/thoraxjnl-2016-209729
      OpenUrlFREE Full Text
    2. ↵
      1. Beasley R,
      2. Chien J,
      3. Douglas J, et al
      . Thoracic society of Australia and New Zealand oxygen guidelines for acute oxygen use in adults: swimming between the flags. Respirology 2015;20:1182–91. doi:10.1111/resp.12620
      OpenUrlCrossRefPubMed
    3. ↵
      1. Piraino T,
      2. Madden M,
      3. Roberts KJ, et al
      . AARC clinical practice guideline: management of adult patients with oxygen in the acute care setting. Respir Care 2022;67:115–28. doi:10.4187/respcare.09294
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Beasley R,
      2. Mackle D,
      3. Young P
      . Oxygen: a new look at an old therapy. J R Soc N Z 2019;49:126–42. doi:10.1080/03036758.2019.1566154
      OpenUrl
    5. ↵
      1. Chu DK,
      2. Kim L-Y,
      3. Young PJ, et al
      . Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis. Lancet 2018;391:1693–705. doi:10.1016/S0140-6736(18)30479-3
      OpenUrlCrossRefPubMed
    6. ↵
      1. Harper JCP,
      2. Semprini R,
      3. Kearns NA, et al
      . Determination of oxygen saturation compared to a prescribed target range using continuous pulse Oximetry in acutely unwell medical patients. BMC Pulm Med 2021;21:332. doi:10.1186/s12890-021-01700-6
    7. ↵
      1. Roca O,
      2. Caritg O,
      3. Santafé M, et al
      . Closed-loop oxygen control improves oxygen therapy in acute hypoxemic respiratory failure patients under high flow nasal oxygen: a randomized cross-over study (the HILOOP study). Crit Care 2022;26:108. doi:10.1186/s13054-022-03970-w
    8. ↵
      1. Ouanes I,
      2. Bouhaouala F,
      3. Maatouk S, et al
      . Automatic oxygen administration and weaning in patients following mechanical ventilation. J Crit Care 2021;61:45–51. doi:10.1016/j.jcrc.2020.10.005
      OpenUrl
    9. ↵
      1. Harper J,
      2. Kearns N,
      3. Bird G, et al
      . Automatic versus manual oxygen titration using a novel nasal high-flow device in medical inpatients with an acute illness: a randomised controlled trial. BMJ Open Respir Res 2021;8:e000843. doi:10.1136/bmjresp-2020-000843
    10. ↵
      1. L’Her E,
      2. Dias P,
      3. Gouillou M, et al
      . Automatic versus manual oxygen administration in the emergency department. Eur Respir J 2017;50:1602552. doi:10.1183/13993003.02552-2016
    11. ↵
      1. L’Her E,
      2. Jaber S,
      3. Verzilli D, et al
      . Automated closed-loop versus standard manual oxygen administration after major abdominal or thoracic surgery: an international multicentre randomised controlled study. Eur Respir J 2021;57:2000182. doi:10.1183/13993003.00182-2020
    12. ↵
      1. Lellouche F,
      2. Bouchard P-A,
      3. Roberge M, et al
      . Automated oxygen titration and weaning with FreeO2 in patients with acute exacerbation of COPD: a pilot randomized trial. Int J Chron Obstruct Pulmon Dis 2016;11:1983–90. doi:10.2147/COPD.S112820
      OpenUrlPubMed
    13. ↵
      1. Hansen EF,
      2. Hove JD,
      3. Bech CS, et al
      . Automated oxygen control with O2Matic® during admission with exacerbation of COPD. Int J Chron Obstruct Pulmon Dis 2018;13:3997–4003. doi:10.2147/COPD.S183762
      OpenUrlPubMed
    14. ↵
      1. Beasley R
      . Swimming between the flags’ with oxygen therapy is getting closer to reality. Respirology 2023;28:960–1. doi:10.1111/resp.14570
      OpenUrl
    15. ↵
      1. Trottier M,
      2. Bouchard P-A,
      3. L’Her E, et al
      . Automated oxygen titration during CPAP and noninvasive ventilation in healthy subjects with induced hypoxemia. Respir Care 2023;68:1553–60. doi:10.4187/respcare.09866
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Munshi L,
      2. Mancebo J,
      3. Brochard LJ, et al
      . Noninvasive respiratory support for adults with acute respiratory failure. N Engl J Med 2022;387:1688–98. doi:10.1056/NEJMra2204556
      OpenUrl
    17. ↵
      1. McKinstry S,
      2. Singer J,
      3. Baarsma JP, et al
      . Nasal high‐flow therapy compared with non‐invasive ventilation in COPD patients with chronic respiratory failure: a randomized controlled cross‐over trial. Respirology 2019;24:1081–7. doi:10.1111/resp.13575
      OpenUrl
    18. ↵
      1. Davidson AC,
      2. Banham S,
      3. Elliott M, et al
      . BTS/ICS guideline for the ventilatory management of acute hypercapnic respiratory failure in adults. Thorax 2016;71 Suppl 2:ii1–35. doi:10.1136/thoraxjnl-2015-208209
      OpenUrlPubMed
    19. ↵
      1. Graham BL,
      2. Steenbruggen I,
      3. Miller MR, et al
      . Standardization of Spirometry 2019 update. An official American Thoracic society and European respiratory society technical statement. Am J Respir Crit Care Med 2019;200:e70–88. doi:10.1164/rccm.201908-1590ST
      OpenUrlCrossRefPubMed
    20. ↵
      1. Pisani L,
      2. Fasano L,
      3. Corcione N, et al
      . Change in pulmonary mechanics and the effect on breathing pattern of high flow oxygen therapy in stable hypercapnic COPD. Thorax 2017;72:373–5. doi:10.1136/thoraxjnl-2016-209673
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Robinson TD,
      2. Freiberg DB,
      3. Regnis JA, et al
      . The role of hypoventilation and ventilation-perfusion redistribution in oxygen-induced hypercapnia during acute exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2000;161:1524–9. doi:10.1164/ajrccm.161.5.9904119
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Parke RL,
      2. Bloch A,
      3. McGuinness SP
      . Effect of very-high-flow nasal therapy on airway pressure and end-expiratory lung impedance in healthy volunteers. Respir Care 2015;60:1397–403. doi:10.4187/respcare.04028
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Groves N,
      2. Tobin A
      . High flow nasal oxygen generates positive airway pressure in adult volunteers. Aust Crit Care 2007;20:126–31. doi:10.1016/j.aucc.2007.08.001
      OpenUrlCrossRefPubMed
    24. ↵
      1. Fraser JF,
      2. Spooner AJ,
      3. Dunster KR, et al
      . Nasal high flow oxygen therapy in patients with COPD reduces respiratory rate and tissue carbon dioxide while increasing tidal and end-expiratory lung volumes: a randomised crossover trial. Thorax 2016;71:759–61. doi:10.1136/thoraxjnl-2015-207962
      OpenUrlAbstract/FREE Full Text
    25. ↵
      1. Brochard L,
      2. Isabey D,
      3. Piquet J, et al
      . Reversal of acute exacerbations of chronic obstructive lung disease by inspiratory assistance with a face mask. N Engl J Med 1990;323:1523–30. doi:10.1056/NEJM199011293232204
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Perrin K,
      2. Wijesinghe M,
      3. Healy B, et al
      . Randomised controlled trial of high concentration versus titrated oxygen therapy in severe exacerbations of asthma. Thorax 2011;66:937–41. doi:10.1136/thx.2010.155259
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Wijesinghe M,
      2. Perrin K,
      3. Healy B, et al
      . Randomized controlled trial of high concentration oxygen in suspected community-acquired pneumonia. J R Soc Med 2012;105:208–16. doi:10.1258/jrsm.2012.110084
      OpenUrlCrossRefPubMed
    28. ↵
      1. Aubier M,
      2. Murciano D,
      3. Milic-Emili J, et al
      . Effects of the administration of O2 on ventilation and blood gases in patients with chronic obstructive pulmonary disease during acute respiratory failure1– 3. Am Rev Respir Dis 1980;122:747–54. doi:10.1164/arrd.1980.122.5.747
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Aubier M,
      2. Murciano D,
      3. Fournier M, et al
      . Central respiratory drive in acute respiratory failure of patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1980;122:191–9. doi:10.1164/arrd.1980.122.2.191
      OpenUrlPubMedWeb of Science
    30. ↵
      1. Sassoon CSH,
      2. Hassell KT,
      3. Mahutte CK
      . Hyperoxic-induced hypercapnia in stable chronic obstructive pulmonary disease1–3. Am Rev Respir Dis 1987;135:907–11. doi:10.1164/arrd.1987.135.4.907
      OpenUrlPubMedWeb of Science
    31. ↵
      1. Crossley DJ,
      2. McGuire GP,
      3. Barrow PM, et al
      . Influence of inspired oxygen concentration on deadspace, respiratory drive, and PaCo2 in intubated patients with chronic obstructive pulmonary disease. Crit Care Med 1997;25:1522–6. doi:10.1097/00003246-199709000-00019
      OpenUrlCrossRefPubMed
    32. ↵
      1. Lellouche F,
      2. L’her E
      . Automated oxygen flow titration to maintain constant oxygenation. Respir Care 2012;57:1254–62. doi:10.4187/respcare.01343
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Rice KL,
      2. Schmidt MF,
      3. Buan JS, et al
      . AccuO2 Oximetry-driven oxygen-conserving device versus fixed-dose oxygen devices in stable COPD patients. Respir Care 2011;56:1901–5.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      1. Vivodtzev I,
      2. L’Her E,
      3. Vottero G, et al
      . Automated O2 titration improves exercise capacity in patients with hypercapnic chronic obstructive pulmonary disease: a randomised controlled cross-over trial. Thorax 2019;74:298–301. doi:10.1136/thoraxjnl-2018-211967
      OpenUrlAbstract/FREE Full Text
    35. ↵
      1. Vézina F-A,
      2. Bouchard P-A,
      3. Breton-Gagnon É, et al
      . Automated O2 titration alone or with high-flow nasal Cannula during walking exercise in chronic lung diseases. Respir Care 2024;69:1–14. doi:10.4187/respcare.10810
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Contributors LK was the guarantor for the overall content of this study, had full access to all the data in this study and took complete responsibility for the integrity of the data and the accuracy of the data analysis including and especially any adverse effects. SK, GB, MB, RS, AE, MW, AS and RB contributed substantially to the study design, data analysis and interpretation, and the writing of the manuscript.

    • Funding Fisher & Paykel Healthcare awarded a research grant to fund this study, were involved in the study design but were not involved in data collection, data analysis, data interpretation, manuscript preparation, or the decision to submit for publication.

    • Competing interests LK, SK, GB, MB, RS, AS, AE and RB report financial support was provided by Fisher & Paykel Healthcare to conduct this study. RB reports that the Institute responsible for this study manuscript has received research funding from Fisher and Paykel Healthcare and the Health Research Council of New Zealand. LK reports a relationship with Fisher & Paykel Healthcare that includes: travel reimbursement. Coauthor RS reports a relationship with Fisher & Paykel Healthcare that includes conference fees. Coauthor GB is now employed by Johnson & Johnson Pacific. Coauthor AS was on a Data Safety Management Committee for a separate study funded by Fisher & Paykel Healthcare.

    • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.