7, 8, 14, 53
Turn the ventilator down at initiation - that is the act that delivers the therapy. Minimise driving pressure and rate, defend PEEP, and escalate the circuit rather than the ventilator when gas exchange deteriorates.
Conventional ventilation
DOI 10.1177/02676591241232270
Lung-protective and ultra-protective ventilation during ECMO
Narrative review. All primary figures quoted here are as reported in this review, not retrieved from the primary papers. Ventilation associations are observational and vulnerable to confounding by severity. No randomised trial of ventilation strategy during VV ECMO exists; the meta-analysis and the randomised trial on proning point in opposite directions.
ELSO guidance: plateau pressure limited to 30 cmH2O (recommended under 25), PEEP at or above 10 cmH2O, respiratory rate 4-15/min, FiO2 as little as possible, and IF GAS EXCHANGE DETERIORATES INCREASE ECLS SUPPORT RATHER THAN VENTILATOR SETTINGS. Schmidt 2019 (n=350): ECMO initiation changed tidal volume 6.4+/-2.0 to 3.7+/-2.0 mL/kg PBW, driving pressure 20+/-7 to 14+/-4 cmH2O, rate 26+/-8 to 14+/-6/min, mechanical power 26.1+/-12.7 to 6.6+/-4.8 J/min; no association between day-2 parameters and mortality. Serpa Neto 2016 pooled IPD (n=545): driving pressure during the first 3 days was the sole independent risk factor for mortality. Schmidt 2015: higher PEEP in early days associated with reduced mortality. Costa 2021: respiratory rate contributes about 25% of the weight of driving pressure to mechanical power, so +4 breaths/min is about +1 cmH2O. 2023 survey of 48 centres: lung rest defined as tidal volume under 5 mL/kg (54%), driving pressure generally under 15 (58% at or below 10), rate under 10/min. Proning during ECMO: Giani 2022 pooled IPD (~900 patients, 5 studies) no survival benefit but proning applied late and in low dose; Papazian 2022 meta-analysis (13 studies, 1836 patients) 28-day survival 74% vs 58%, RR 1.31 (1.21-1.41); PRONECMO RCT (Schmidt 2023, 94% COVID) no benefit on any endpoint, no severe adverse events; 74% of surveyed centres prone during VV ECMO. NMBA used by 93% of respondents in 2023 but withdrawn as soon as possible, versus 41% in a 2019 prospective study; ATS conditional recommendation in favour, ESICM no recommendation.
Adults on VV and VA ECMO