Safety and Outcomes of Prolonged Usual Care Prone Position Mechanical Ventilation to Treat Acute Coronavirus Disease 2019 Hypoxemic Respiratory Failure

Objectives: Prone position ventilation is a potentially life-saving ancillary intervention but is not widely adopted for coronavirus disease 2019 or acute respiratory distress syndrome from other causes. Implementation of lung-protective ventilation including prone positioning for coronavirus disease 2019 acute respiratory distress syndrome is limited by isolation precautions and personal protective equipment scarcity. We sought to determine the safety and associated clinical outcomes for coronavirus disease 2019 acute respiratory distress syndrome treated with prolonged prone position ventilation without daily repositioning.

Design: Retrospective single-center study.

Setting: Community academic medical ICU.

Patients: Sequential mechanically ventilated patients with coronavirus disease 2019 acute respiratory distress syndrome.

Interventions: Lung-protective ventilation and prolonged protocolized prone position ventilation without daily supine repositioning. Supine repositioning was performed only when Fio2 less than 60% with positive end-expiratory pressure less than 10 cm H2O for greater than or equal to 4 hours.

Measurements and main results: Primary safety outcome: proportion with pressure wounds by Grades (0-4). Secondary outcomes: hospital survival, length of stay, rates of facial and limb edema, hospital-acquired infections, device displacement, and measures of lung mechanics and oxygenation. Eighty-seven coronavirus disease 2019 patients were mechanically ventilated. Sixty-one were treated with prone position ventilation, whereas 26 did not meet criteria. Forty-two survived (68.9%). Median (interquartile range) time from intubation to prone position ventilation was 0.28 d (0.11-0.80 d). Total prone position ventilation duration was 4.87 d (2.08-9.97 d). Prone position ventilation was applied for 30.3% (18.2-42.2%) of the first 28 days. Pao2:Fio2 diverged significantly by day 3 between survivors 147 (108-164) and nonsurvivors 107 (85-146), mean difference -9.632 (95% CI, -48.3 to 0.0; p = 0·05). Age, driving pressure, day 1, and day 3 Pao2:Fio2 were predictive of time to death. Thirty-eight (71.7%) developed ventral pressure wounds that were associated with prone position ventilation duration and day 3 Sequential Organ Failure Assessment. Limb weakness occurred in 58 (95.1%) with brachial plexus palsies in five (8.2%). Hospital-acquired infections other than central line-associated blood stream infections were infrequent.

Conclusions: Prolonged prone position ventilation was feasible and relatively safe with implications for wider adoption in treating critically ill coronavirus disease 2019 patients and acute respiratory distress syndrome of other etiologies.