Prone positioning does not affect survival in patients with ARDS

Context

The exact mechanism by which oxygenation is improved in patients ventilated in a prone position (compared to those ventilated in supine position) is not known but may be due to reductions in ventilation/perfusion (VQ) mismatching and chest wall compliance. Improvement in oxygenation is noted in about 60% of patients; significant numbers sustain improvement after being returned to a supine position. Careful positioning usually requires three to five people. Complications are rare, although hemodynamic instability (1.1% per prone cycle), accidental extubation (0.4%), central line dislodgement (0.4%), pressure ulcers (15%) (see Additional information [1]) have all been reported.

Significant findings

Neither intention to treat nor per-protocol analysis revealed significant differences in the primary outcomes. The prone group had a larger improvement in ratios of partial pressure of arterial oxygen to the fraction of inspired oxygen (PaO₂/FiO₂) at day 10 (63 versus 45 [P = 0.02]) and slightly higher tidal volumes (10.7 versus 10.7 [P = 0.03]). The prone group had a higher incidence of new pressure sores - 2.7% versus 1.9% (P = 0.004). There were no significant differences in endotracheal tube or venous access displacement. However, prone positioning resulted in increased requirements for sedation (55%) and neuromuscular blockade (27.7%), and more episodes of transient airway obstruction (39%) and hypotension (12%). In a post-hoc analysis, a subgroup of patients with the lowest PaO₂/FiO₂ ratio (<88) in the prone group had a lower 10-day mortality, but this did not persist to discharge from the ICU.

Comments

The majority of patients with acute respiratory distress syndrome (ARDS) die not from hypoxemia but from multiple-organ failure (see Additional information [2]). Thus, improvement in surrogate outcomes (such as PaO₂/FiO₂ ratio) with prone positioning may be misleading. Minimizing tidal volumes to prevent ventilator associated lung injury at the price of accepting lower physiologic values of PaO₂and pH has lead to improved survival (see Additional information [3]). The patients in this study were ventilated with larger tidal volumes than currently recommended. Although generally safe (in a research setting), routine use of prone positioning cannot be recommended as yet because appropriate timing and duration of prone positioning remain unknown.

Methods

A total of 304 patients from 28 ICUs met the criteria (see Additional information [4]) for ARDS or acute lung injury. They were assigned randomly to either the prone (n = 152) or supine group (n = 152). Patients in the prone group were kept prone for at least six hours per day for 10 days. Physicians used standardized ventilator settings (see Additional information [5]). Primary endpoints were mortality at 10 days, ICU discharge and six months after randomization; secondary endpoints were oxygenation and organ dysfunction at 10 days.