1. Dallas 2015 TFQO: Tony Scott #138 EVREV 1: Tony Scott #138 EVREV 2: Anthony Seto # Taskforce: ACS Supplemental Oxygen in ACS 887

2. Dallas 2015 COI Disclosure (specific to this systematic review) Tony Scott COI#138 Commercial/industry none Potential intellectual conflicts None Anthony Seto COI# Commercial/industry None Potential intellectual conflicts None

3. Dallas 2015 2010 CoSTR Oxygen has been used in the treatment of myocardial infarction and acute coronary syndromes for many years. Evidence in support of this approach is primarily derived from animal models, in which the administration of oxygen during and/or after experimental coronary artery occlusion reduces the extent of myocardial necrosis in some (Maroko 1975, Ribeiro 1979, Kelly 1995, Ishikawa 1986), but not all studies (Malm 1977, Weisse 1982). Few human studies conducted in small number of patients suffering a myocardial infarction reported diminished cardiac output during oxygen therapy (Thomas 1965, Kenmure 1968, Foster 1969). One human case study found improvement in ST changes with the use of oxygen (Madias 1976). Rawles and Kenmure conducted a blinded randomized controlled trial in 157 patients suffering myocardial infarction (oxygen versus room air inhalation). Aspartate aminotransferase levels, as surrogate for infarct size, were higher in the oxygen group (P<0.05). There were 11.3% deaths in the oxygen group and 4% in the air group, relative risk of death 2.9 (95% CI 0.8-10.3, P=0.08). Ventricular tachycardia occurred in 13.8% of the oxygen group and 6.5 % of the air groups, relative risk 2.1 (0.8-5.8, P=0.13). The authors concluded that there was suggestive evidence of a deleterious effect of oxygen. This study lacked statistical power to detect clinically important outcomes. Wilson and Channer conducted a parallel-group, non-blinded controlled study of oxygen therapy. Fifty patients with myocardial infarction treated with streptokinase were randomized to oxygen or room air for 24h. The study was not designed to compare outcomes measured by infarct size or mortality. Ventricular tachycardia occurred in 23% of the oxygen group and 25% of the air group (P=NS). Also there was no significant difference in opiate use. Severe hypoxemia occurred in 8 patients (1 in the oxygen group versus 7 in the air group, P<0.05) There is insufficient evidence to support the use of oxygen in the treatment of uncomplicated myocardial infarction (Nicholson 2004, Wijesinghe 2009). The routine use of oxygen in this situation may increase infarct size and possibly increase the risk of mortality. There is a need for randomized controlled trials of the use of oxygen therapy in uncomplicated myocardial infarction that are sufficiently powered and performed in the current reperfusion era.

4. Dallas 2015 C2015 PICO P: Among adult patients with suspected ACS and normal oxygen saturation in any setting (prehospital, emergency or in hospital) I: Withholding oxygen C: Routine supplemental oxygen O: Death, infarct size, chest pain resolution, ECG resolution

5. Dallas 2015 Inclusion/Exclusion & Articles Found Inclusions Withholding supplemental oxygen in normoxic patients with acute coronary syndromes Exclusions Intracoronary oxygen, hyperbaric oxygen, reviews, animal studies, abstract only Initial search 1648 articles, 17 full text reviewed, none included included 4 RCTs identified from systematic reviews 0 non-RCTs 1648 excluded Revised search strategy currently being tested

6. Dallas 2015 2015 Proposed Treatment Recommendations We suggest against routine oxygen supplementation in comparison to air in normoxic patients with acute coronary syndromes (weak recommendation, low quality of evidence). In making this recommendation we place a higher value on avoiding possible harm when the evidence available suggests no mortality benefit and possible harm in providing routine oxygen supplementation.

7. Dallas 2015 Risk of Bias in studies

8. Dallas 2015 Evidence profile table

9. Dallas 2015 Forest Plots

10. Dallas 2015 Forest Plots

11. Dallas 2015 Proposed Consensus on Science statements For the critical outcome of mortality we have identified very low quality evidence (downgraded for indirectness, heterogeneity and bias) from three RCTs (Rawles 1976 p1121, Ukholkina 2005 p59, Ranchord 2012 p168 ) enrolling 430 patients showing no benefit (OR 0.46 95% CI 0.16 to 1.29) when oxygen is withheld. For the important outcome of infarct size we have identified low quality evidence (downgraded for bias, inconsistency, indirectness and imprecision) from two RCTs (Rawles 1976 p1121, Ranchord 2012 p168) enrolling 274 patients showing mild reduction in infarct size when oxygen is withheld. Data from a 3 rd RCT suggesting increased infarct size could not be used formally due to incomplete reporting and invalidated methods (Ukholkina 2005 p59). The trial data generated for infarct size is too heterogeneous to enable combined assessment

12. Dallas 2015 Proposed Consensus on Science statements For the important outcome of chest pain resolution we have identified very low quality evidence (downgraded for bias, inconsistency, indirectness and imprecision) from two RCTs (Rawles 1976 p1121, Wilson 1997 p657) enrolling 199 patients showing no difference when oxygen is withheld. For the important outcome of ECG resolution no evidence has been identified in RCTs Important data from the AVOID trial has been presented and awaits publication. There was no difference in mortality, however there was evidence of a significant reduction in infarct size and reinfarction when supplemental oxygen was withheld.

13. Dallas 2015 Draft Treatment Recommendations We suggest against routine oxygen supplementation in comparison to air in normoxic patients with acute coronary syndromes (weak recommendation, low quality of evidence). In making this recommendation we place a higher value on avoiding possible harm when the evidence available suggests no mortality benefit and possible harm in providing routine oxygen supplementation.

14. Dallas 2015 Knowledge Gaps Appropriately powered trials in the modern revascularization era are required. Several are currently either underway or awaiting publication