Table 4

Baseline variables which were examined to determine if there was a differential benefit from ambulatory oxygen compared with ambulatory air

(Subgroup) VariableSubgroupsP ValuesOxygen–air
T×V*VTEstimate§95% CI
None0.3120.775(−0.736 to 2.285)
DesaturationNo**0.3930.7060.3360.744(−0.780 to 2.268)
Severity of airflow obstructionFEV1% <500.8380.3320.3780.681(−0.841 to 2.203)
Exercise response to oxygen6MWD‡‡ <54 m0.7680.5620.1681.126(−0.480 to 2.731)
n=127≥54 m
Volume response to hyperoxiaΔIC§§ <10%0.2240.1150.2470.887(−0.621 to 2.395)
Severe breathlessnessBDI score¶¶≤40.1530.7590.3550.715(−0.809 to 2.238)
GenderFemale0.3170.4420.3380.738(−0.778 to 2.254)
All of the above; main effects only≥0.1710.2011.075(−0.581 to 2.730)
  • Data represent the analysis of covariance results, with week 12 values as the response variable, the corresponding value at baseline as the covariate and each of the subgroup variables as an explanatory factor, in addition to treatment (air or oxygen).

  • * Treatment (oxygen or air) by (subgroup) variable interaction.

  • (Subgroup) variable.

  • Treatment: oxygen or air.

  • § Estimate of the mean difference in dyspnoea scores between subjects on oxygen and those on air, after allowing for dyspnoea scores at baseline (as a covariate) and the (subgroup) variable.

  • 95% CI associated with the estimate (in previous column).

  • ** SpO2 at the conclusion of a 6 min walk test breathing air. No=SpO2 >88%, Yes=SpO2 ≤88%.

  • †† Sample size used for the analysis; one extreme outlier was omitted from all analyses while the level of the (subgroup) variable was unknown in some cases.

  • ‡‡ Increase in 6 min walk distance (6MWD) breathing 6 l/min O2 compared with air beyond the reported minimally important difference.

  • §§ Increase in percentage predicted inspiratory capacity (ΔIC) while breathing 44% oxygen compared with 21% oxygen.

  • ¶¶ Dyspnoea dichotomised as severe or not according to the Baseline Dyspnoea Index (BDI) focal score.

  • FEV1, forced expiratory volume in 1 s.