Open Access Te Herenga Waka-Victoria University of Wellington
thesis_access.pdf (11.86 MB)

Closed-Loop Oxygen Control

Download (11.86 MB)
posted on 2023-07-10, 22:55 authored by James Harper

Guidelines recommend that oxygen should be titrated to achieve a target oxygen saturation (SpO2 ) range in acutely unwell patients, a concept colloquially known as “swimming between the flags”. The limited available evidence suggests the current practice of manual oxygen titration commonly does not achieve the prescribed SpO2 target range, resulting in exposure to inadequate and/or excessive concentrations of oxygen, and potential risk of harm. The aim of the research presented in this thesis is to investigate a novel method of oxygen delivery, using nasal high-flow (NHF) with automatic oxygen titration, to achieve SpO 2 within the prescribed target range.

An initial feasibility study was undertaken, whereby the closed-loop algorithm used for automatic oxygen titration was tested in participants with resting hypoxaemia. This study also investigated physiological responses to changes in flow and oxygen concentration, enabling the algorithm to be developed and optimised. The percentage of time spent with SpO 2 within a target range using closed-loop control improved during the 12 study visits from 65% at the first visit to 90% at the last visit.

The validated version of the algorithm was evaluated in a controlled laboratory-based setting. A randomised crossover study was undertaken involving 42 participants with chronic lung disease who desaturated during exertion. The main aim was to determine the ability of automatic oxygen titration to maintain SpO2 within a target range during a six-minute walk test (6MWT) and subsequent 10-minute recovery period. The comparator interventions were room air and a fixed concentration of oxygen, both delivered using NHF. During the 6MWT, the proportion of time spent with SpO2 in target range using automatic titration was greater than room air; difference 26.0% (95%CI 17.7 to 34.2), P<0.001, and not differ significantly from the fixed concentration of oxygen; -8.2% (95%CI -16.5 to 0.1), P=0.052. In comparison, during the rest period the proportion of time spent in range using automatic titration was greater than fixed concentration of oxygen; difference 19.3% (95%CI 8.9 to 29.7), P<0.001, and did not differ significantly from room air; -9.3% (95%CI -19.7 to 1.0), P=0.077. These results demonstrated proof of concept efficacy of the system in responding to changes in SpO2 outside of a target range.

The efficacy of manual oxygen titration in terms of maintenance of SpO2 within a prescribed target range was assessed in two studies. A retrospective audit of 62 medical patients in whom a prescribed target SpO2 range was present, demonstrated the mean (SD) proportion of SpO2 measurements recorded in the inpatient observation chart which were within range while receiving oxygen was 56.2% (30.6). The mean percentage of measurements above range was 33.5% (33.8) and 10.3% (17.8) below range. Patients with a reduced hypercapnic target range (88-92% or lower) had a higher probability of having SpO2 above range; multivariate odds ratio 5.34 (95%CI 1.65 to 17.3), P=0.006.

A non-interventional study using continuous pulse oximetry for 24-hours in 80 acutely unwell medical patients receiving oxygen was undertaken in order to assess the time exposure to SpO2 within, above and below range, and to investigate which factors influence this. The mean (SD) proportion of time spent in target range was 55.6% (23.6); this was lower in those with a reduced hypercapnic target range compared to those with a range of 92-96%, difference - 13.1% (95%CI -3.0 to -23.2), P=0.012. The proportion of time spent above range was 16.2% (22.9); this was higher in those with a reduced hypercapnic range, difference 21.6% (95%CI 31.4 to 12), P<0.001. The proportion of time below range was 28.4% (25.2), with no difference between target ranges.

A randomised controlled trial was undertaken in this same medical inpatient population comparing NHF with automatic oxygen titration to NHF with manual oxygen titration for a period of 24-hours. 20 participants were included in the analysis. Automatic oxygen titration resulted in a median (IQR) 96.2% (95.2 to 97.8) of time within the target range compared to 71% (59.4 to 88.3) with manual titration; difference 24.2% (95%CI 7.9 to 35), P<0.001. This marked difference confirmed the efficacy of automatic oxygen titration in acutely unwell medical patients.

This series of clinical studies has evaluated NHF with automatic oxygen titration from initial feasibility, to proof-of-concept efficacy and has ultimately demonstrated superiority compared to current practice in the acute inpatient setting. The additional studies undertaken have demonstrated a need for this technology based on the difficulty in achieving a target SpO 2 range according to the current practice of manual oxygen titration. This body of work paves the way for a paradigm shift in the delivery of oxygen for acutely unwell patients and presents a means to achieve the main objective of oxygen therapy as described by the concept of “swimming between the flags”. Further studies in different patient populations are now required to confirm these findings and larger studies are needed to assess the impact of automatic oxygen titration on clinical outcomes.


Copyright Date



Te Herenga Waka—Victoria University of Wellington

Rights License

Author Retains All Rights

Degree Discipline

Biological Sciences

Degree Grantor

Te Herenga Waka—Victoria University of Wellington

Degree Level


Degree Name

Doctor of Philosophy

Victoria University of Wellington Item Type

Awarded Doctoral Thesis



Victoria University of Wellington School

School of Biological Sciences


Ackerley, David; Beasley, Richard