52 in 52 – #26: The EXACT Trial

Welcome back to the “52 in 52” series. This collection of posts features recently published must-know articles. Post #26 looks at oxygen targets in the post-ROSC patient after OHCA.

Author: Dan O’Conor, MD (Resident, Emergency Medicine Physician, San Antonio, TX) and Brannon Inman, MD (Chief Resident, Emergency Medicine Physician, San Antonio, TX) // Reviewed by: Alex Koyfman, MD (@EMHighAK); and Brit Long, MD (@long_brit)

Study design:

• A multi-center parallel-group block-randomized controlled trial
  • Patients and outcome assessors blinded

PICO:

Population:

• Included patients in 2 EMS systems and 15 hospitals in Victoria and South Australia, Australia from December 11, 2017- August 11, 2020
• Included adult patients (above age 18) with OHCA from presumed cardiac cause with subsequent ROSC
• Additional inclusion criteria:
  • unconscious after ROSC
  • advanced airway (endotracheal tube or supraglottic airway)
  • transport planned to a participating hospital
  • maintained SpO2 >95% while receiving 10L/min supplemental O2 or FiO2 100%
• Exclusion criteria:
  • Pregnancy
  • Evidence of non-cardiac cause of arrest

Intervention:

• Oxygen restriction: Supplemental oxygen titration prior to ICU arrival to achieve SpO2 90-94%

Comparator:

• Liberal oxygen: Supplemental oxygen titration prior to ICU arrival to achieve SpO2 98-100%

Outcome:

• No significant difference in the primary outcome (survival to hospital discharge)
• Statistically significant increased risk of hypoxic episodes (SpO2 <90%) in the intervention group

Randomization:

• Treatment arms were fairly well-balanced.

Take away:

• This was a negative trial, although it was limited by early termination prior to target enrollment due to the COVID pandemic.
• There were increased rates of sustained hypoxia in the intervention group, although this “adverse effect” is an oxygen range immediately adjacent to the target range.
• Strengths:
  • This study includes patient-oriented outcomes and outcomes that are directly related to the hypothesized mechanism (CPC neurologic assessment scale as a secondary outcome when the premise is that standard care causes reperfusion brain injury).
• This study contrasts with prior preliminary studies suggesting a benefit of conservative oxygen strategies.
• Authors note that the precise oxygen titration was difficult with the equipment available pre-hospital.
  • Mean SpO2 in the intervention group was 97% (range 94-99%) at ED arrival despite a goal of 90-94%.
  • This decreases the likelihood that if a difference exists between groups that it could be detected by this study, although there was a noticeable difference in PaO2 at ED arrival.
    • Median 95mmHg in the intervention group vs 130mmHg in the standard group without any PaO2 <60 in the intervention group which suggests that hyperoxia was in fact mitigated in the intervention group without causing hypoxia.
  • The only subgroup in which survival varied was patients with or without bystander CPR. This may support the theory that patients with shorter downtime/hypoxia are at less risk of hypoxia and therefore do not stand to benefit as significantly from measures aimed to reduce reperfusion injury. This should be viewed as hypothesis-generating at this stage.

My take:

I will not intentionally target hyperoxia, and it is not clear whether 90-94% is inferior or superior to 98-100% based on this trial. There is likely a greater number of desaturation episodes when targeting 90-94%, but it is not clear this is associated with patient harm.

Reference:

1. Bernard SA, Bray JE, Smith K, et al. Effect of Lower vs Higher Oxygen Saturation Targets on Survival to Hospital Discharge Among Patients Resuscitated After Out-of-Hospital Cardiac Arrest: The EXACT Randomized Clinical Trial. JAMA. 2022;328(18):1818-1826. doi:10.1001/jama.2022.17701