Memsorb™, a novel CO2 removal device part II: in vivo performance with the Zeus IE®.

Memsorb™ (DMF Medical, Halifax, Canada) is a novel device based upon membrane oxygenator technology designed to eliminate CO2 from exhaled gas when using a circle anesthesia circuit. Exhaled gases pass through semipermeable hollow fibers and sweep gas flowing through these fibers creates a diffusion gradient for CO2 removal. In vivo Memsorb™ performance was tested during target-controlled closed-circuit anesthesia (TCCCA) with desflurane in O2/air using a Zeus IE® anesthesia workstation (Dräger, Lübeck, Germany). Clinical care protocols for using this novel device were guided by in vitro performance results from a prior study (submitted simultaneously). After IRB approval, written informed consent was obtained from 10 ASA PS I-III patients undergoing robot-assisted radical prostatectomy. TCCCA targets were 39% inspired O2 concentration (FIO2) and 5.0% end-expired desflurane concentration (FETdes). Minute ventilation (MV) was adjusted to maintain 4.5-6.0% FETCO2. The O2/air (40% O2) sweep flow into the Memsorb™ was manually adjusted in an attempt to keep inspired CO2 concentration (FICO2) ≤ 0.8%. The following data were collected: FIO2, FETdes, FICO2, FETCO2, MV, fresh gas flow (FGF, O2 and air), sweep flow, and cumulative desflurane usage (Vdes). Vdes of the Zeus IE®-Memsorb™ combination was compared with historical Vdes observed in a previous study when soda lime (DrägerSorb 800 +) was used. Results are reported as median and inter-quartiles. A combination of manually adjusting sweep flow (26 [21,27] L/min) and MV sufficed to maintain FICO2 ≤ 0.8% and FETCO2 ≤ 6.0%, except in one patient in whom the target Zeus IE® FGF had to be increased to 0.7 L/min for 6 min. FIO2 and FETdes were maintained close to their targets. Zeus IE® FGF after 5 min was 0 [0,0] mL/min. Average Vdes after 50 min was higher with Memsorb™ (20.3 mL) compared to historical soda lime canister data (12.3 mL). During target-controlled closed-circuit anesthesia in patients undergoing robot-assisted radical prostatectomy, the Memsorb™ maintained FICO2 ≤ 0.8% and FETCO2 ≤ 6.0%, and FIO2 remained close to target. Modest amounts of desflurane were lost with the use of the Memsorb™. The need for adjustments of sweep flow, minute ventilation, and occasionally Zeus IE® FGF indicates that the Memsorb™ system should preferentially be integrated into an automated closed-loop system.

Memsorb™, a novel CO2 removal device part I: in vitro performance with the Zeus IE®.

Soda lime-based CO2 absorbents are safe, but not ideal for reasons of ecology, economy, and dust formation. The Memsorb™ is a novel CO2 removal device that uses cardiopulmonary bypass oxygenator technology instead: a sweep gas passes through semipermeable hollow fibers, adding or removing gas from the circle breathing system. We studied the in vitro performance of a prototype Memsorb™ used with a Zeus IE® anesthesia machine when administering sevoflurane and desflurane in O2/air mixtures. The Zeus IE® equipped with Memsorb™ ventilated a 2L breathing bag with a CO2 inflow port in its tip. CO2 kinetics were studied by using different combinations of CO2 inflow (VCO2), Memsorb™ sweep gas flow, and Zeus IE® fresh gas flow (FGF) and ventilator settings. More specifically, it was determined under what circumstances the inspired CO2 concentration (FICO2) could be kept < 0.5%. O2 kinetics were studied by measuring the inspired O2 concentration (FIO2) resulting from different combinations of Memsorb™ sweep gas flow and O2 concentrations, and Zeus IE® FGFs and O2 concentrations. Memsorb™'s sevoflurane and desflurane waste was determined by measuring their injection rates during target-controlled closed-circuit anesthesia (TCCCA), and were compared to historical controls when using a soda lime absorbent (Draegersorb 800+) under identical conditions. With 160 mL/min VCO2 and 5 L/min minute ventilation (MV), lowering the sweep gas flow at any fixed Zeus IE® FGF increased FICO2 in a non-linear manner. Sweep gas flow adjustments kept FICO2 < 0.5% over the entire Zeus IE® FGF range tested with VCO2 up to 280 mL/min; tidal volume and respiratory rate affected the required sweep gas flow. At 10 L/min MV and low FGF (< 1.5 L/min), even a maximum sweep flow of 43 L/min was unable to keep FICO2 ≤ 0.5%. When the O2 concentration in the Zeus IE® FGF and the Memsorb™ sweep gas flow differed, FIO2 drifted towards the sweep gas O2 concentration, and more so as FGF was lowered; this effect was absent once FGF > minute ventilation. During sevoflurane and desflurane TCCCA, the Zeus IE® FGF remained zero while agent usage per % end-expired agent increased with increasing end-expired target agent concentrations and with a higher target FIO2. Agent waste during target-controlled delivery was higher with Memsorb™ than with the soda lime product, with the difference remaining almost constant over the FGF range studied. With a 5 L/min MV, Memsorb™ successfully removes CO2 with inflow rates up to 240 mL/min if an FICO2 of 0.5% is accepted, but at 10 L/min MV and low FGF (< 1.5 L/min), even a maximum sweep flow of 43 L/min was unable to keep FICO2 ≤ 0.5%. To avoid FIO2 deviating substantially from the O2 concentration in the fresh gas, the O2 concentration in the fresh gas and sweep gas should match. Compared to the use of Ca(OH)2 based CO2 absorbent, inhaled agent waste is increased. The device is most likely to find its use integrated in closed loop systems.

In vitro efficiency of 16 different Ca(OH)2 based CO2 absorbent brands.

Data directly comparing CO2 absorbents tested in identical and clinically relevant conditions are scarce or non-existent. We therefore tested and compared the efficiency of 16 different brands of Ca(OH)2 based CO2 absorbents used as loose fill or a cartridge in a refillable canister under identical low flow conditions. CO2 absorbents efficiency was tested by flowing 160 mL/min CO2 into the tip of a 2 L balloon that was ventilated with an ADU anesthesia machine (GE, Madison, WI, USA) with a tidal volume of 500 mL and a respiratory rate of 10/min while running an O2/air FGF of 300 mL/min. After the 1020 mL refillable container was filled with a known volume of CO2 absorbent (derived from weighing the initial canister content and the product's density), the time for the inspired CO2 concentration (FICO2) to rise to 0.5% was measured. This test was repeated 4 times for each product. Because the two SpiraLith Ca® products (one with and one without indicator) are delivered as a cartridge, they had to be tested using their proprietary canister. The time (min) for FICO2 to reach 0.5% was normalized to 100 mL of product, and defined as the efficiency, which was compared amongst the different brands using ANOVA. Efficiency ranged from 50 to 100 min per 100 mL of product, and increased with increasing NaOH content (a catalyst), the exception being SpiraLith Ca® cartridge with color indicator (performing as well as the most efficient granular products) and the SpiraLith Ca® cartridge without color indicator (outperforming all others). Results indicated a spherical or bullet shape is less efficient in absorbing CO2 than broken fragments or cylinders, which in turn is less efficient than a hemispherical (disc) shape, which is in turn less efficient than a solid cartridge with a molded channel geometry. The efficiency of Ca(OH)2 based CO2 absorbent differs up to 100% on a volume basis. Macroscopic arrangement (cylindrical wrap with preformed channels versus granules), chemical composition (NaOH content), and granular shape all affect efficiency per volume of product. The data can be used to compare costs of the different products.