Reference | Type of study | Patients included | Prone position characteristics | Main results | Adverse events |
---|---|---|---|---|---|
Kipping et al. [19] | Retrospective cohort | 12 | Duration of prone sessions: 8 h Number of prone sessions: median of 6 (IQR 4–8) Total number of prone sessions: 74 | 58% reported an improvement in the PaO2/FiO2 ratio > 20% No change in MAP, HR, but mPAP significantly increased during proning and decreased after proning Dose of norepinephrine could also be decreased | One lost NGT Bleeding from ECMO cannulation sites (11/74) or tracheal tube (10/74) or central venous lines (8/74) or chest tubes (10/74) One endotracheal tube obstruction One pulmonary embolism Drop in SpO2 > 2% (10/74) One temporary reduced blood flow of ECMO Hemodynamic instability (7/74) Temporary bradycardia (3/74) |
Masuda et al. [20] | Cohort | 5 | Duration of prone sessions: mean 15.3 ± 0.5 h Number of prone sessions: mean 1.8 ± 0.8 | Oxygenation improvement: PaO2/FiO2 ratio in supine 143 ± 38 mmHg vs. prone 263 ± 99 mmHg | None |
Guervilly et al. [21] | Prospective cohort | 15 | Initiation after a median of 9 (IQR 5–10) days on ECMO Duration of prone sessions: 12 h Number of prone sessions: 1.4 per patient Total number of prone sessions: 21 | Oxygenation improvement: PaO2/FiO2 ratio increased from 103 (78–135) vs. 160 (96–215); P = 0.007 The oxygenation improvement persisted after returning the patients to the supine position No changes in PaCO2 and Crs were observed | No major adverse events Variations in ECMO flow were small (1.6 ± 4% compared to baseline) Two patients required crystalloid infusions of 500 ml for MAP < 65 mmHg during proning One pneumothorax occurring during proning was diagnosed and drained only after returning to supine position |
Kimmoun et al. [18] | Retrospective cohort | 17 | Initiation after a median of 6 (4–12) days on ECMO Duration of prone sessions: 24 h Total number of prone sessions: 27 | Oxygenation improvement: PaO2/FiO2 ratio in supine 111 (IQR 84–128) mmHg vs. at the end of prone session 173 (120–203) mmHg Oxygenation improvement occurred more frequently in patients who were proned after 7 days of ECMO therapy Improvement in Crs: From 18 (12–36) to 32 (15–36) ml/ cmH2O 24 h after the return to supine position, tidal volume was increased from 3.0 (2.2–4.0) to 3.7(2.8–5.0) ml/kg No correlation was observed between the oxygenation improvement and the amount of non-aerated lung tissue in the CT scan | One membrane thrombosis, one drop in ECMO blood flow |
Lucchini et al. [22] | Retrospective cohort | 14 | Duration of prone sessions: median 8 h (IQR 6–10) Number of prone sessions: (median 1—IQR 1–1.5) Total number of prone sessions: 45 | Oxygenation improvement: PaO2/FiO2 ratio in supine 123 (IQR 82–135) mmHg and at the end of prone session 149 (90–186) mmHg This improvement was not maintained when the patient was turned to supine 113 (74–182) No significant hemodynamic variations (HR, SAP, PAPm, CO, PWP, SvO2) | None |
Rilinger et al. [26] | Retrospective cohort propensity score matched | 38 | Initiation after a median of 1.7 (0.5–5.0) days on ECMO Duration of prone sessions: median 19.5 (IQR 16.8–20.8) hours Number of prone sessions: 2 (1–3) | No difference in hospital survival (36.8% vs. 36.8%, P = 1.0) No difference in ECMO weaning rate (47.4% vs. 44.7%, P = 0.82) Hospital survival was superior in the subgroup of patients treated with early proning (< 17 h) as compared to late or no proning (81.8% vs. 33.3%, P = 0.02) 60-day mortality was 18% for the early proning and 65% for the late and no proning group, respectively (P = 0.027) Survival rate of early proning was higher compared to late proning or no prone (81.8% vs. 18.5% and 36.7%, P < 0.001 and P = 0.003, respectively) | No relevant complications |
Franchineau et al. [25] | Prospective cohort | 21 | Duration of prone sessions: 16 h | Static Crs during proning increased from 23 (17–29) to 27 (20–37) ml/cmH2O (P < 0.01) 13 (62%) patients increased their static Crs by 3 ml/cmH2O after proning on ECMO (mechanical responders) EELI was redistributed from ventral to dorsal regions during proning Optimal PEEP determined by EIT was lower in prone position (14 (12–16) vs. 10 (8–14) cmH2O | None |
Garcia et al. [23] | Retrospective cohort | 14 (SARS- CoV-2) | Duration of prone sessions: median 16 h (IQR 15–17) Total number of prone sessions: 24 | Oxygenation improvement: PaO2/FiO2 ratio in supine 84 (IQR 73–108) vs 112 (83–157) after proning The median PaO2/FiO2 ratio improvement after proning was 28% [2–36]. 62.5% high responders (increase PaO2/ FiO2 ratio > 20%), 16.7% moderate- responders (increase PaO2/FiO2 < 20%), and 20.8% non- responders (decrease PaO2/FiO2) Patients in the prone ECMO group were less likely to be weaned from ECMO, and 28-day mortality rate was significantly higher | Three minor hemorrhages at site of cannula insertion Three moderate flow drops of VV-ECMO that required fluid resuscitation |
Giani et al. [24] | Multicenter retrospective cohort propensity score matched | 240 patients (66 matched pairs) | Initiation after a median of 4 (IQR 2–7) days on ECMO Duration of proning: mean 15 (12–18) h Total number of prone sessions: 326 | Improvement in oxygenation, intrapulmonary shunt fraction and static Crs that persisted after supination Minor differences in hemodynamics (mPAPm and PWP were slightly higher during proning and HR was lower) Lower hospital mortality in proned patients (OR = 0.50, 95%CI: 0.29–0.87) PS matched cohort: proned patients had a lower mortality (30% vs. 53%, P = 0.0241) proned patients had a longer duration of ECMO (16 vs10 days, P = 0.0344) | No major complication 6% minor complications Six procedures aborted due to respiratory or hemodynamic instability during prone positioning |