Key Point:

• A decrease in PaCO2 after six hours of prone positioning predicts improved survival in patients with ALI or ARDS.

MILAN—Years of clinical experience have proved that prone positioning during mechanical ventilation increases oxygenation in patients with acute lung injury (ALI) or acute respiratory distress syndrome (ARDS). But does pronation improve clinical outcomes in these patients?

In some patients, yes—but not necessarily because of its positive effect on oxygen status. A recent study of more than 200 patients with ALI or ARDS found that carbon dioxide clearance may be more important in determining clinical outcomes.[1] Gattinoni and colleagues reported that among patients placed in the prone position, 28-day survival correlated inversely with changes in arterial carbon dioxide tension (PaCO2).[1] No association between survival and arterial oxygen tension (Pao2) was found.

Lead author Luciano Gattinoni, MD, believes that the significance of carbon dioxide clearance has been overlooked. “PaCO2 clearance is usually neglected or under-considered. Carbon dioxide changes, more so than oxygen changes, relate to the anatomical status of the lung,” he explained. Dr. Gattinoni is a Professor of Anesthesia and Intensive Care at the University of Milan.

Other experts concur. “We have become so fixated on oxygenation as an indicator of improvement in patients with ALI and ARDS that we have forgotten the importance of carbon dioxide removal in these patients,” stressed Thomas Bein, MD, in an interview. Dr. Bein, a Professor of Anesthesia at the University Hospital Regensburg in Germany, wrote an editorial commenting on the Gattinoni study.[2]

Dr. Gattinoni and colleagues retrospectively analyzed data from their 2001 trial of prone positioning, as well as from a similar unpublished pilot study. The same protocol had been used for the patients in both studies—they had been placed in the prone position for at least six hours daily for 10 days; no changes were made to the ventilator settings. Data from 209 patients were analyzed.

Patients were categorized as PaCO2 responders if they had experienced a drop in that variable of 1 mm Hg or more after the first six hours of pronation. In contrast, three different criteria were used to define PaO2 responders. At the end of the first six hours of pronation, the ratio of PaO2 to fraction of inspired oxygen (FiO2) had to have risen by:
• At least 1 mm Hg.
• 20 mm Hg or more.
• Any amount above the median increase in the Pao2/Fio2 ratio, which was 58 mm Hg.

More than 85% of the patients experienced a marked increase in oxygenation following six hours of pronation; the mean increase in the PaO2/FiO2 ratio was 70 mm Hg. Yet, no differences in survival between PaO2 responders and nonresponders could be detected, regardless of which definition of a PaO2 responder was used.

In contrast, only 45% of the patients were PaCO2 responders. Furthermore, the mean changes in PaCO2 were small: +0.4 mm Hg in the group as a whole and –6.0 mm Hg in the PaCO2 responders. However, the difference in 28-day survival between PaCO2 responders and nonresponders was striking: 65% versus 48%. When changes in PaCO2 were analyzed as a continuous variable, survival was inversely associated with the size of the decrease in PaCO2 among responders, and among nonresponders, it correlated directly with the size of the increase in that measurement.

Why does the postpronation change in PaCO2 predict outcome in patients with ALI or ARDS? The researchers believe it is not the increase in carbon dioxide clearance per se that is responsible; rather, the change in PaCO2 reflects differences in the underlying pathophysiology. “When PaCO2 decreases [after the patient is placed] in the prone position, the prevalent underlying pathology is lung collapse—like a wet sponge, the lung collapses under its own weight,” suggested Dr. Gattinoni. “When PaCO2 increases, the prevalent phenomenon is likely pulmonary consolidation with lung blood flow redistribution.”

In the first setting, turning the patient over reduces the weight and alleviates the collapse; this may contribute to improved survival. In the second setting, a change from the supine to prone position would not alter the consolidation and thus would not be expected to affect clinical outcomes.

However, both PaCO2 responders and nonresponders may experience a marked increase in oxygenation after several hours of pronation. In the PaCO2 responders, the most probable cause of the rise in Pao2 is the recruitment to aeration of previously collapsed regions; in the PaCO2 nonresponders, it most likely results from diversion of blood flow from nonaerated to aerated regions.

“Intensive care physicians know from experience that the prone position is effective in ALI and ARDS,” said Dr. Bein. “But this study … returns us to a place from which we can recommend the prone position as an evidence-based measure.” Dr. Gattinoni’s 2001 trial was a setback that made such a recommendation more difficult, Dr. Bein explained, because it failed to associate pronation with improved survival in ALI and ARDS.

—Timothy Begany

References
1. Gattinoni L, Vagginelli F, Carlesso E, et al. Decrease in PaCO2 with prone position is predictive of improved outcome in acute respiratory distress syndrome. Crit Care Med. 2003;31:2727-2733.
2. Bein T. Prone position, carbon dioxide elimination, and survival: a turn for the better? Crit Care Med. 2003; 31:2804-2805.

Return to table of contents