Table 1.
Relevant studies comparing conservative vs. liberal oxygen strategies in acute illness.
| Study | Methods | Findings | Comment | Suggestions |
|---|---|---|---|---|
| Chu et al. (18) | Meta-analysis (up to Oct 2017) of trials comparing liberal vs. conservative oxygen strategies in a variety of conditions(n = 16,037) | “hyperoxemia” carries an increased 1-year mortality risk [HR 1.11 (95% CI 1.00–1.24), p = 0.05) | Heterogenous studies, examining oxygen therapy mainly in ischemic conditions. Unreliable in relation to oxygen targets in acute nonischemic illness. |
Optimum target saturations might become unfavorable above SpO2 of 94–96% |
| Girardis et al. (22) | Single-centered, open-labeled RCT Conservative (SpO2 of 94–95%) vs. liberal (SPO2 of 97–100%) in ICU admissions of any cause (n = 434) |
Improved mortality in conservative group (SpO2 of 94–98%): RR 0.57 (95% CI 0.37–0.9, p = 0.01) | Liberal group had significantly more medical problems at enrolment. | Target sats of 94–98% |
| Panwar et al. (23) | Pilot multicenter RCT SpO2 of 88–92% vs. 96%in patients requiring MV of any cause (n = 103) |
No difference in mortality or length of ICU stay | 21% of the conservative group had COPD vs. only 10% of the liberal. Actual comparison was SpO2 of 93.4 vs. 97% | Larger trial needed |
| Schernthaner et al. (24) | Retrospective observational studies comparing arterial blood gases and mortality in pulmonary edema and heart failure (n = 475) | Increased mortality in patients with pneumonia with arterial PO2 of 150 vs. 117 mmHG (HR = 1.02; 95% CI: 1–1.4, p = 0.02) | Pneumonia related Nontrial data Potentially measuring true hyperoxemia effects. |
Optimal PO2 calculated as 98 mmHg (or ~SpO2 of 97.3%) |
| van den Boom et al. (17) | Multicenter, retrospective observational analysis of SpO2 and mortality in ICU patients of any cause (n = 35,000) | Increased mortality with mean SpO2 92 vs. 96% [OR, 3.2 (2.9–3.5)]. Increased mortality with mean SpO2 100 vs. 96% [OR 1.6 (1.5–1.6)] (n = 26,723). | Large patient numbers utilizing shared datasets. Not a clinical trial. |
Optimum target SpO2 is 94–98% |
| IICU-ROX Investigators the Australian New Zealand Intensive Care Society Clinical Trials Group et al. (20) | Multicenter, RCT SpO2 of 92–96% vs. normal in patients requiring MV of any cause (n = 1,000) | No difference in mortality or length of ICU stay | Actual Comparison was mean SpO2 of 95–96% vs. 96–97% | None made |
| Barrot et al. (21) | Multicenter, RCT SpO2 of 88–92% vs. >96%in patients with ARDS of any cause (n = 205) | Study halted due to safety. Significantly higher mortality in the conservative oxygen group. Ninety-day mortality was 44% in conservative arm vs. 30.4% in liberal arm. |
Actual Comparison was mean SpO2 of 92–93% vs. 95–97% | None made |
RCT, randomized controlled trial; SpO2, oxygen saturations; HR, hazard ratio; RR, relative risk; OR, odds ratio (adjusted); ICU, intensive care unit; MV, mechanical ventilation.