Demand oxygen delivery systems (DODS) enable prolongation of liquid oxygen cylinder life compared to continuous oxygen flow (CONT) use. Evidence is lacking, however, regarding their efficacy. This study investigated the literature comparing liquid-based CONT to DODS in patients with chronic obstructive pulmonary disease (COPD). Four electronic databases were searched from 1980 until January 2018. Primary outcomes were oxyhaemoglobin saturation (SpO2 %) at rest and during exercise and exercise performance. Risk of bias was evaluated using the Cochrane tool. Data were analysed via meta-analysis where possible using the generic inverse variance method in Revman 5.3 or narrative synthesis. Ten crossover trials involving 152 patients with moderate to severe COPD (forced expiratory volume in 1 s (FEV1 ) range: 27-42% predicted) were included. There was a large degree of uncertainty regarding potential bias related to allocation concealment and blinding for all studies. Data from three studies (n = 44) showed no significant differences between DODS and CONT in terms of SpO2 % at rest -0.2% (95% CI: -0.5% to 0.1%) or during exercise -0.3% (95% CI: -2.1% to 1.5%). The pooled mean difference of two studies (n = 56) in 6-min walk distance was 5.7 m (95% CI: -14.4 to 25.8 m). Findings were consistent between the meta-analysis and narrative synthesis. These findings from a limited number of studies suggest oxygen delivery via DODS or CONT confers similar effects in terms of SpO2 % or exercise performance in patients with COPD. However, as DODS devices use various specifications that may yield large intra-individual differences, individual SpO2 % testing appears advisable for those considering DODS use.
Keywords: chronic obstrutive pulmonary disease; hypoxemia; oxygen device; oxygen therapy.
© 2018 Asian Pacific Society of Respirology.