19, 22, 23
The largest dataset behind Chapter 19's argument that reperfusion is an intervention rather than an event. At cannulation the oxygen fraction is 1.0 and the sweep is high, both correctly; ten minutes later both are probably wrong and nobody owns the decision. Name the person responsible for the first right-radial arterial gas and the time it will be taken.
Normoxia and normocarbia
[VERIFICATION REQUIRED] — the bibliographic index attaches a preprint identifier (DOI 10.1101/2022.03.10.22272203) to the journal record. The journal DOI was not established and is not guessed.
Peri-cannulation arterial oxygen and carbon dioxide tension, categorised as hypoxia, normoxia, mild/moderate/severe hyperoxia and hypocarbia, normocarbia, mild/severe hypercarbia
Registry data with only two gas values per patient and multiple imputation for missing values. Unavoidable confounding: hyperoxia on ECMO may mark a chaotic resuscitation rather than cause harm. Association only. The effect sizes are modest but consistent across four separate outcomes.
Composite acute brain injury in 488 (16%): 7% ischaemic stroke, 3% intracranial haemorrhage. On-ECMO moderate hyperoxia (PaO2 200–299) aOR 1.42 (1.02–1.97) and severe hyperoxia (PaO2 at or above 300) aOR 1.59 (1.20–2.10) for composite acute brain injury. Severe hyperoxia: ischaemic stroke aOR 1.63 (1.11–2.40), intracranial haemorrhage aOR 1.92 (1.08–3.40), in-hospital mortality aOR 1.58 (1.21–2.06). Mild hypercarbia pre-ECMO protective for composite injury aOR 0.61 (0.44–0.84) and ischaemic stroke aOR 0.56 (0.35–0.89).
3,125 adults aged 18 or over undergoing ECPR in the ELSO Registry, January 2009 to December 2020. Median age 58; 69% male. Two blood-gas values per patient (6 hours pre-cannulation, 24 hours post). Missing values handled by multiple imputation.