The use of venovenous extracorporeal membrane oxygenation (VV ECMO) for severe acute respiratory distress syndrome (ARDS) has risen substantially, and current guidelines recommend VV ECMO in selected patients to maintain adequate gas exchange while limiting ventilator-induced lung injury. Despite this, several critical components of ECMO management remain poorly standardised, particularly the weaning process.
Weaning from VV ECMO is inherently complex, involving multiple interacting physiological and clinical variables. Evidence-based protocols are lacking, and current practice is largely guided by expert opinion. Important aspects such as the optimal timing of sweep gas-off trials (SGOT), criteria for decannulation, and ventilatory strategies during the transition from ECMO remain uncertain. In response, an international survey was developed to characterise real-world VV ECMO weaning practices worldwide.
The descriptive, semi-structured, anonymous international survey was designed, formally reviewed, and endorsed by the Acute Respiratory Failure Section of the European Society of Intensive Care Medicine (ESICM), and distributed through ESICM communication channels, partner societies, social media, and the April 2025 Extracorporeal Life Support Organization newsletter.
The questionnaire consisted of four sections: (1) respondent background and experience, (2) screening criteria for VV ECMO weaning, (3) SGOT practices, and (4) ventilation management during weaning. Each item was rated on a scale from 1 (not important) to 5 (very important). For descriptive analysis, responses were dichotomised into important or very important (scores 4–5) versus less important (scores 1–3).
A total of 303 responses were received from 187 centres across 46 countries. Most responses came from Europe (54%) and North America (21%). The majority of respondents were critical care physicians (83%), and 75% reported more than five years of ECMO management experience. The median annual ECMO case volume was 20 (10–30) VV ECMO cases and 20 (8–40) VA ECMO cases.
The perceived importance of different variables used to assess readiness for ECMO weaning was identified. Parameters related to arterial blood gases were rated as most important, including PaO₂ (94.5%), PaCO₂ (96.2%), and pH (91.2%). Measures of respiratory mechanics were also highly valued, including tidal volume (89%), respiratory system compliance (88%), and driving pressure (88%). Positive end-expiratory pressure (PEEP) was considered important by 76% of respondents, and the transition to assisted ventilation by 64%. Imaging was regarded as less influential: 51% rated chest x-ray and 39% rated lung ultrasound as important or very important.
Most centres (60%) reported using a protocolised SGOT, although only 34% applied standardised trial durations. Among those using standardised trials, the median SGOT duration was 12 (6–24) hours. Where trials were not standardised, they most commonly lasted between 6 and 24 hours (65%).
The perceived importance of different variables for evaluating SGOT success was also evaluated. Arterial gas exchange again dominated, with PaCO₂ (98%), PaO₂ (94%), and pH (93%) most frequently rated as important or very important. Respiratory mechanics were also considered relevant, including respiratory rate (88%), tidal volume (87%), driving pressure (82%), and respiratory system compliance (80%). Respiratory drive, mixed venous oxygen saturation, and systolic arterial pressure were less commonly regarded as important. Only 30% of respondents considered oesophageal pressure monitoring to be relevant for assessing readiness to wean from ECMO.
Extubation before ECMO weaning was reported as uncommon by 72% of respondents. Among patients who remained intubated (91%), controlled ventilation (49%) and assisted ventilation (42%) were used with similar frequency during the weaning phase. Approximately 20% of participants reported routinely performing early tracheostomy, defined as a tracheostomy performed within the first few days of ECMO to reduce sedative requirements.
Overall, the WEAN-ECMO survey demonstrated substantial variability in VV ECMO weaning practices. Clinicians relied predominantly on gas exchange variables, with objective measures of respiratory mechanics considered relevant but less influential. Oesophageal pressure monitoring was rarely used. Although most centres reported using a protocolised SGOT, there was wide variation in trial duration and evaluation criteria. The lack of a formal definition of “protocolised” in the survey may have resulted in some over-reporting and contributed to the observed heterogeneity. Ventilatory strategies during SGOT were also inconsistent. Weaning failure was generally attributed to persistent gas-exchange abnormalities, respiratory distress, or clinical instability. This variability in weaning mirrors the broader heterogeneity seen in ECMO patient selection and management.
Source: Intensive Care Medicine
Image Credit: iStock
