Advice

Clinical and technical evidence

Clinical and technical evidence

A literature search was carried out for this briefing in accordance with the interim process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.

Published evidence

There are 6 studies summarised in this briefing, including 552 people, selected as the most relevant and best quality evidence relating to the technology. Four studies were randomised controlled trials and 2 were prospective single-centre crossover studies.

An economic study was also located (Poder et al. 2018) and is commented on in the resource consequences section.

The clinical evidence and its strengths and limitations are summarised in the overall assessment of the evidence.

Overall assessment of the evidence

The evidence base for the technology is of moderate methodological quality. All studies had standard care comparators. Two used multi-site recruitment and both had good size populations. One study reported upon its use in babies and children. All studies suggest that FreeO2 may increase the time in the target SpO2 range. None of the reported studies are based in the NHS. Although the company reports that the device can be used to treat acute respiratory distress syndrome (ARDs) from COVID‑19, none of the summarised evidence includes this indication.

Further evidence would benefit from use of the device in UK‑based settings across both chronic obstructive pulmonary disease (COPD) and ARD indications and in all ages.

Ouanes et al. 2021

Intervention and comparator

FreeO2 and constant flow modes.

Key outcomes

Time spent within target SpO2 range was significantly higher with FreeO2 mode compared with constant O2 flow mode (86.92% [77.11% to 92.39%] compared with 43.17% [5.08% to 75.37%]; p<0.001). Time with hyperoxia was lower with FreeO2 mode: 8.68% (2.96% to 15.59%) compared with 38.28% (2.02% to 86.34%). Times with hypoxaemia and with severe desaturation were similar. At the end of FreeO2 mode, O2 flow was lower than 1 litre/min in 28 people (54.9%), with a median of 0.99 litre/min.

Strengths and limitations

This was a pilot crossover design focused on physiological parameters without collecting patient-centred outcomes. People were not recruited consecutively because of availability of 1 FreeO2 device. Some results were reported to not be interpretable because of loss of the SpO2 signal.

Roué et al. 2021

Intervention and comparator

FreeO2 and standard manual O2.

Key outcomes

Time spent within the SpO2 predefined target range was significantly increased in the FreeO2 group (94.6% plus or minus 6%, compared with 76.3% plus or minus 22%), showing a difference of 18.4 (confidence interval [CI] 10.1 to 26.7). Secondary measure of time spent with severe desaturation did not significantly differ between groups (0.04% plus or minus 0.17%, compared with 0.15% plus or minus 0.62% in the control group). The children in the FreeO2 group spent significantly less time with hypoxemia, especially in the older children group (0.8% in the FreeO2 group and 22% in the control group).

Strengths and limitations

The FreeO2 device used in the pilot study was a prototype using an algorithm that was subsequently modified in the latest version of the device. Authors recognised the limitation of the low mean number of changes in oxygen flow seen in the control group, which may be affected by the patient-nurse ratio. The device was only used in the first 24 hours of hospitalisation and for a mean duration of 210 minutes. One author is a cofounder and shareholder of the research and development company for FreeO2.

L'Her et al. 2021

Intervention and comparator

Automated closed-loop oxygen administration using FreeO2 (n=103) and standard closed-loop oxygen administration (n=95).

Key outcomes

The primary outcome was the percentage of time within the oxygenated range during a 3‑day time frame, which was shown to be 31.9% increased time in the range in the automated group (95% CI -42.8% to 59.2%). The secondary outcomes results included: periods of hypoxaemia reduced in the automated group by -10.2% (95% CI -13.9 to -6.6%), periods of hyperoxaemia reduced in the automated group by -22.0% (95% CI -27.6% to -16.4%).

Strengths and limitations

Investigators were not blinded. Monitoring in the standard group was done using the FreeO2 device, which may have resulted in more frequent monitoring than standard care and have reduced the benefits seen in automated administration. One author of L'Her is a cofounder and shareholder of the research and development company for FreeO2.

L'Her et al. 2017

Intervention and comparator

FreeO2 (n=93) and manual oxygen titration (n=94).

Key outcomes

Time within the SpO2 target was higher under automated titration (81% plus or minus 21% compared with 51% plus or minus 30%, p<0.001). Automated titration significantly reduced time with hypoxaemia (3% plus or minus 9% compared with 5% plus or minus 12%, p=0.04) and hyperoxia under O2 (4% plus or minus 9% compared with 22% plus or minus 30%, p<0.001). Oxygen could be weaned at the end of the study in 14.1% compared with 4.3% of people in the automated and manual titration group, respectively (p<0.001). O2 duration during the hospital stay was significantly reduced (5.6 days plus or minus 5.4 compared with 7.1 days plus or minus 6.3, p=0.002).

Strengths and limitations

Randomisation was sealed to the intervention and intention-to-treat analysis was used. However, the length of follow-up period (3 hours) may not have captured outcome improvement. One author (L'Her) is a cofounder and shareholder of the research and development company for FreeO2.

Lellouche et al. 2016

Intervention and comparator

FreeO2 and manual oxygen titration.

Key outcomes

Significantly higher percentage of time spent in target SpO2 and reduced time in severe desaturation and hyperoxia. Time from study inclusion to hospital discharge was reduced but not significantly (5.8 days plus or minus 4.4 with FreeO2 and 8.4 days plus or minus 6.0 with usual oxygen administration; p=0.051).

Strengths and limitations

Blinding of all investigators was not possible because of the practical set up of the system, but steps were taken to blind those that could be. The study has a small sample size and may not be generalisable to a wider range of disease severity. Two authors are coinventors of FreeO2. Two authors participate in Onnovair, a company that owns shares in OxyNov.

Schneeberger et al. 2021

Intervention and comparator

FreeO2 and constant flow modes.

Key outcomes

This study found significantly and clinically relevant improvements in walking endurance time, with 68% of people walking for longer in the automated titration group. Reasons for stopping the endurance shuttle walk tests was significantly different between constant titration and automated titration support (p=0.001). Dyspnoea was the main reason (in 70% of people) for stopping the test with constant flow modes, compared with 48% stopping because of breathlessness with automatic oxygen titration. People in the study also reported to prefer the automatic O2 system.

Strengths and limitations

The authors highlighted that 14 people reached the maximum exercise duration and the effect size may have been different with extended time periods. The study represents the immediate effects of O2 therapy and might not reflect longer usage scenarios. The authors recommend that future studies should consider medium and long-term effects of using automatically titrating oxygen system. No conflicts were declared.

Sustainability

The device is reported to optimise oxygen use and may reduce overall oxygen use as shown in Poder et al. 2018 and L'Her et al. 2021.

Recent and ongoing studies