Hypoxemia in trauma patients receiving two different oxygen strategies: a TRAUMOX2 substudy

Table 2 Pulse oximetry for trauma patients allocated to either a restrictive or a liberal oxygen strategy

	Restrictive oxygen group (N = 30)	Liberal oxygen group (N = 30)	Difference (95% CI)	P value
SpO₂ data, median (IQR), hours	6.8 (6.3–7.3)	7.1 (6.3–7.4)
SpO₂ < 90% for > 5 min	2 (7%)	1 (3%)	3% (-10–16%)	1.00^‡
SpO₂ < 90% for > 2 min	3 (10%)	3 (10%)	0% (-16–16%)	1.00^‡
SpO₂ <85% for > 5 min	1 (3%)	1 (3%)	0% (-12–12%)	1.00^‡
SpO₂ <90% for any duration	25 (83%)	19 (63%)	19% (-3–40%)
SpO₂ <85% for any duration	18 (60%)	15 (50%)	9% (-15–34%)
SpO₂ < 90% cumulative time, sec^∗	58 (6–152)	13 (0–139)	12 (-4–57)	0.16^§
SpO₂ < 90% duration episodes, sec^†	9 (4–15)	10 (5–22)	-2 (-4–-1)	0.008^§

Abbreviations: CI, confidence interval; IQR, interquartile range; sec, seconds; SpO₂, arterial oxygen saturation measured by pulse oximetry
Notes: For categorical outcomes, the 95% confidence interval for difference in proportions was calculated by the modified Wald method. For continuous outcomes, the median difference was computed as the Hodges-Lehmann estimate
^∗The median of cumulative time per patient
^†The median duration of all episodes per group
^‡Fisher’s exact test
^§Mann-Whitney U test

ISSN: 1757-7241