Mechanical Ventilation in the Obese Patient: Compliance, Pleural Pressure, and Driving Pressure

Obesity is increasingly common in Western societies (1). When critically ill, obese patients present many management challenges, especially during mechanical ventilation (2). As a consequence of the large abdominal and chest wall loads on the diaphragm, they have more atelectasis and hypoxemia and require higher pleural pressure (Ppl) and airway pressure to maintain adequate oxygen saturation as measured by pulse oximetry (Sp_O2). These higher pressures have the potential to decrease Q̇. This can negate the benefit of an increase in Sp_O2 and result in no change or even a decrease in O₂ delivery (DO₂), which ultimately is what matters for tissues. There is little information on airway pressure management in obese patients because they usually are left out of clinical trials. Accordingly, in this issue of the Journal, to evaluate the hemodynamic consequences of higher levels of airway pressure in obese patients with acute respiratory distress syndrome (ARDS), De Santis Santiago and colleagues (pp. 575–584) (3) performed clinical and animal studies to determine if higher positive end-expiratory pressure (PEEP) can improve gas exchange without compromising hemodynamics.

In a crossover design with 19 obese patients who had an average body mass index of 57 ± 12 kg/m², they compared the hemodynamic effects of PEEP based on the standard ARDS network PEEP table (4) versus higher PEEP determined by a lung recruitment procedure and PEEP titrated to respiratory system compliance as in ART (Alveolar Recruitment for Acute Respiratory Distress) (5). In a subset, they also compared changes in regional lung ventilation and perfusion by electrical impedance tomography in these patients and selected nonobese patients from ART (5).

There was no evidence of hemodynamic compromise with the higher PEEP in the obese subjects, nor echocardiographic evidence of right ventricle dysfunction, although the measurements were of limited sensitivity. In the subset with electrical impedance tomography studies, the lung recruitment strategy produced more homogeneous ventilation and reduced lung collapse by 31% without causing overdistention. Respiratory system compliance increased by 24%, driving pressure, which is the difference between the plateau of inspiratory pressure and PEEP, decreased by 30%, and Pa_O2/Fi_O2 markedly increased. In patients without obesity, overdistention was more common in the nondependent regions and lung perfusion was highly heterogeneous. It was considered too invasive to measure Q̇ and DO₂, but unfortunately, these are the key variables needed for interpreting the results.

It was in the animal study that the hemodynamic benefit of higher PEEP is evident. The authors compared PEEP 7 versus 19 cm H₂O in normal swine and swine with obesity and ARDS simulated by placing a weight on the abdomen and lung lavage. It is worth noting some design deficiencies. A weight on the abdomen produces a homogeneous increase in abdominal pressure and misses the effects of intraabdominal fat acting primarily on the dorsal diaphragm and the chest wall load. However, these issues likely give quantitative differences but do not compromise the qualitative response. It also was unfortunate that the authors only compared the equivalent of animals with obesity and ARDS with normal swine rather than a third group with ARDS and no obesity. Without it, the hemodynamic effect of ARDS cannot be fully separated from that of obesity. Ppl was measured with esophageal balloons (6). This allowed vascular pressures to be presented as the transmural pressure (intravascular minus the outside pleural pressure) as well as pressures relative to atmosphere, which is necessary to understand the relationship of the heart to the rest of body. Most importantly, they also measured Q̇ and calculated DO₂.

Differences in the hemodynamic responses to the high PEEP between the two groups were striking. Control swine had a marked fall in mean arterial pressure, a rise in pulmonary arterial pressure (PAP), and minimal changes in the transmural central venous pressure (CVP) and wedge pressure. Most significantly, Q̇ and DO₂ fell by more than 30%. In contrast, in the obese lung injury swine, PAP fell and there was no change in transmural CVP and wedge pressure and only a modest 12% fall in Q̇; DO₂ actually rose. The rise in DO₂ with a fall in Q̇ was at first hard to explain, as was the marked rise in mixed venous saturation from a mean of 52–75% with no change in V̇o₂. Working backward from the O₂ extraction fraction, it is apparent that this occurred because of a marked increase in arterial Sp_O2 from the 65% range before the recruitment to close to 100% after.

What accounts for the marked difference in Q̇ response in the obese versus nonobese condition with high PEEP? Mechanical ventilation decreases Q̇ either by altering venous return to the heart by increasing CVP relative to atmospheric pressure (and not the transmural CVP) or by loading the RV. In the healthy swine, high PEEP increased CVP by 6 mm Hg relative to atmosphere and, by decreasing venous return, likely was the primary cause of the fall in Q̇. There was a small increase in transmural CVP and no change in transmural RV pressure, suggesting only a small inspiratory increase in RV afterload from an increase in transpulmonary pressure (1). Interpretation of the RV load is difficult. A decrease in venous return and Q̇ decrease PAP, whereas increased RV load raises PAP, which also lowers Q̇ and changes cardiac filling pressures.

In the swine with obesity and ARDS, the recruitment maneuver markedly improved lung compliance so that driving pressure decreased and there only was a modest increase in inspiratory transpulmonary pressure. As a result, there was a smaller fall in venous return and Q̇. The recruitment maneuver also resulted in a striking reduction in the inspiratory load on the RV as evidenced by the fall in pulmonary artery pressure and transmural RV systolic pressure.

The major determinant of the inspiratory load on the RV is not the actual Ppl but rather driving pressure. In the obese patients with ARDs, driving pressure dramatically decreased from 13 ± 4 to 9 ± 2 cm H₂O because of the improved respiratory system compliance following recruitment of collapsed lung and better distribution of blood flow. This reinforces the observation that driving pressure is a key variable to follow during ventilator management (7). Based on this study, the argument can be made that a lower driving pressure is not only lung protective but also an important factor for cardiac protection. A second component was the large improvement in Sp_O2 from improved V/Q matching.

Two other observations are worth commenting on. By improving V/Q matching, the rise in Sp_O2 increased DO₂ and more than compensated for the small fall in Q̇. The message is that all parts of the DO₂ equation need to be considered when managing patients. The second is historical. In the 1990s, there was a lot of discussion about supply-dependent V̇o₂ (8). Calculated V̇o₂ in all animal groups were strikingly similar, indicating that this value most often is regulated by the underlying metabolic activity and not DO₂.

As a cautionary note, although lung recruitment improved DO₂, the same protocol in ART (5) showed net harm. We suggest that it may be safer to use an escalating rather than a deescalating PEEP trial to identify best total thoracic compliance. In this approach, PEEP is increased with a fixed inspiratory pressure until Vt decreases. The PEEP below this value is then used. This likely gives a PEEP value that is lower than that determined by an initial recruitment and deescalation of PEEP because of the hysteresis between inspiration and expiration the curves, but it is safer and likely still adequate for the hemodynamic benefit.

In conclusion, higher levels of PEEP in obese patients with ARDS reduces harmful heart–lung interactions. The primary benefit derives from improving respiratory system compliance, which then allows for a lower driving pressure to ventilate the lung and consequently less compromise of RV function. This further emphasizes the clinical value of following driving pressure.