Auto-positive end-expiratory pressure during tracheal gas insufflation: Testing a hypothetical model

Objective: The major benefit of tracheal gas insufflation (TGI) is an incre ase in CO2 elimination efficiency by removal of CO2 from the anatomical dea dspace. In conjunction with mechanical ventilation, TGI may also alter vari ables that affect CO2 elimination, such as minute ventilation and peak airw ay pressure (peak Paw) and cause the development of auto-positive end-expir atory pressure (auto-PEEP). We tested the hypothesis that TGI-induced auto- PEEP alters ventilatory variables, We predicted that TGI-induced auto-PEEP offsets the beneficial effects of TGI on CO2 elimination and that keeping t otal PEEP (ventilator PEEP + auto-PEEP) constant enhances the CO2 eliminati on efficiency afforded by TGI. Design: Prospective study of two series of patients with acute respiratory distress syndrome receiving mechanical ventilation. Setting: Intensive care units at a university medical center. Patients: Each series consisted of eight sequential hypercapnic patients. Interventions: In series 1, we examined the effect of continuous TGI at 0 a nd 10 L/min on PaCO2, without compensating for the development of auto-PEEP . In series 2, we examined this same effect of continuous TGI while reducin g ventilator PEEP to keep total PEEP constant. TGI-induced auto-PEEP was ca lculated based on dynamic compliance measurements during zero TGI flow cond itions (Delta V/Delta P) after averaging the two baseline values for peak P aw and tidal volume and assuming compliance did not change between the zero TGI and TGI flow conditions (Delta V-TGI/Delta P-TGI). Measurements and Main Results: In series 1, total PEEP increased from 13.2 +/- 3.2 cm H2O to 17.8 +/- 3.5 cm H2O without compensation for auto-PEEP (p = .01). PaCO2 decreased (p = .03) from 56.2 +/- 10.6 mm Hg (zero TGI) to 5 2.9 +/- 9.3 mm Hg (TGI at 10 L/min), a 6% decrement. In series 2, total PEE P was unchanged (p = NS). Pace, decreased (p = .03) from 59.5 +/- 10.4 mm H g (zero TGI) to 52.2 +/- 8.3 mm Hg (161 at 10 L/min), a 12% decrement There was no significant change in PaO2; there were no untoward hemodynamic effe cts in either series. Conclusions: These data are consistent with the hypothesis that mechanical ventilation + TGI causes an increase in auto-PEEP that can blunt CO2 elimin ation. In addition to the ventilator modifications necessary to keep ventil atory variables constant when TGI is used, it is also necessary to reduce v entilator PEEP to keep total PEEP constant and further enhance CO2 eliminat ion efficiency.