Targeting arterial partial pressure of carbon dioxide in acute respiratory distress syndrome patients using extracorporeal carbon dioxide removal.

BACKGROUND: A retrospective analysis of SUPERNOVA trial data showed that reductions in tidal volume to ultraprotective levels without significant increases in arterial partial pressure of carbon dioxide (PaCO2 ) for critically ill, mechanically ventilated patients with acute respiratory distress syndrome (ARDS) depends on the rate of extracorporeal carbon dioxide removal (ECCO2 R).

METHODS: We used a whole-body mathematical model of acid-base balance to quantify the effect of altering carbon dioxide (CO2 ) removal rates using different ECCO2 R devices to achieve target PaCO2 levels in ARDS patients. Specifically, we predicted the effect of using a new, larger surface area PrismaLung+ device instead of the original PrismaLung device on the results from two multicenter clinical studies in critically ill, mechanically ventilated ARDS patients.

RESULTS: After calibrating model parameters to the clinical study data using the PrismaLung device, model predictions determined optimal extracorporeal blood flow rates for the PrismaLung+ and mechanical ventilation frequencies to obtain target PaCO2 levels of 45 and 50 mm Hg in mild and moderate ARDS patients treated at a tidal volume of 3.98 ml/kg predicted body weight (PW). Comparable model predictions showed that reductions in tidal volumes below 6 ml/kg PBW may be difficult for acidotic highly severe ARDS patients with acute kidney injury and high CO2 production rates using a PrismaLung+ device in-series with a continuous venovenous hemofiltration device.

CONCLUSIONS: The described model provides guidance on achieving target PaCO2 levels in mechanically ventilated ARDS patients using protective and ultraprotective tidal volumes when increasing CO2 removal rates from ECCO2 R devices.