Since Murray et al. (1) introduced the new CO2 absorbent Amsorb® (Armstrong Medical Ltd., Coleraine, Northern Ireland) in 1999, it has been welcomed as an absorbent because it attenuates the degradation of volatile anesthetics as compared with other absorbents. Amsorb prevents the degradation of sevoflurane to Compound A, and the degradation of desflurane, enflurane, or isoflurane to carbon monoxide (CO) (1–4). Although the difference between the CO2 absorption capacity of Amsorb and that of soda lime is only 10%–15% in the in vitro study of Murray et al. (1), we often experienced a faster than expected exhaustion of Amsorb. Several in vitro studies (5–8) consistently reported that the difference between the CO2 absorption capacity of Amsorb and that of soda lime was 30%–50%. These clinical experiences and the confusing results among the in vitro studies prompted us to investigate the CO2 absorption capacity of Amsorb under clinical conditions, because there is no information regarding the longevity of Amsorb with low-flow (1 L/min) anesthesia in vivo, which might serve as a good reference value for clinical anesthetists. Thus, the purpose of this study was to compare the CO2 absorption capacity of Amsorb with that of soda lime, i.e., Medisorb® (Datex-Ohmeda, Bromma, Sweden) and Sodasorb® (W.R. Grace, Lexington, MA), in a rebreathing system with clinical low-flow (1 L/min) anesthesia.
The study was conducted at the Self Defense Force Central Hospital in Tokyo, Japan, and was approved by the hospital ethics committee. An informed consent form was signed by each patient before participation in the study. The subjects were 69 patients undergoing general anesthesia for various surgeries. Two different CO2 absorbent canisters were investigated.
Study 1: Small Canister
The anesthesia machine used was an ADU (Datex-Ohmeda, Helsinki, Finland). This machine has one compact 750-mL canister that can be changed during ventilation. CO2 elimination (V̇Eco2) by the patient and minute volume were measured via the airway by using indirect calorimetry (CS/3; Datex-Ohmeda, Helsinki, Finland), in which the CO2 analyzers were calibrated with 95% oxygen and 5% CO2. The mea-sured gas volumes were corrected every minute to standard temperature and pressure. The accuracy of the V̇Eco2 value is within ±10% less than a fraction of inspired oxygen of 0.65. Nitrous oxide and inhaled anesthetics were excluded because they can affect the performance of the CS/3 metabolic monitor. Anesthesia was induced by the administration of 10 L/min oxygen followed by 2–2.5 mg/kg propofol, 100–200 μg fentanyl, and 0.05–0.10 mg/kg vecuronium IV. After tracheal intubation, the flow rates of oxygen and air were set to 3 and 7 L/min, respectively. After 3 min, the fresh gas flow rate was reduced to 1 L/min: the flow rates of oxygen and air were set at 400 and 600 mL/min, respectively. The flow rates of oxygen and air were adjusted to maintain the inspiratory oxygen concentration at approximately 40%. Anesthesia was maintained with a propofol (6–10 mg/kg) and fentanyl (50–100 μg/kg) infusion. Ventilation was controlled with a tidal volume of 8–10 mL/kg, and the ventilatory rate was adjusted to maintain an end-tidal CO2 of 35–40 mm Hg. After skin closure, anesthetic administration was discontinued, and the fresh gas inflow rate was changed to 10 L/min of oxygen. Once the patients opened their eyes and took a deep breath on verbal command, the endotracheal tube was removed.
Amsorb, which does not contain potassium hydroxide (KOH) or sodium hydroxide (NaOH); Medisorb, which contains 2% NaOH and trace amounts of KOH (0.003%); or Sodasorb, which contains 3% KOH and 2% NaOH, were used as the CO2 absorbents. The type of CO2 absorbent to be used was selected among Amsorb, Medisorb, and Sodasorb by drawing lots, and the CO2 absorbent was then used repeatedly until exhaustion. We prepared 700 mL of fresh CO2 absorbent by using a graduated cylinder and then weighed the absorbent before use. The absorbent was then placed into the canister. The CO2 absorbent was not changed for each patient but was left in the canister until the completion of the first anesthesia, during which CO2 rebreathing was observed, at which time the CO2 absorbent was replaced with one of the other CO2 absorbents, selected by drawing lots. Rebreathing during general anesthesia was defined as an increased inspired pressure of CO2 ≥2 mm Hg, which decreased after the increase in the total gas flow and reached 2 mm Hg again when the total gas flow was reduced to 1 L/min (9). This procedure was performed 18 times to obtain data for six trials each for Amsorb, Medisorb, and Sodasorb in a randomized manner.
Study 2: Large Canister
The study with the large canister was performed after the completion of Study 1. The anesthesia machine used was an Aestiva 3000 (Datex-Ohmeda, Helsinki, Finland). This machine has two (upper and lower) large 1350-mL canisters. We prepared 700 mL of fresh CO2 absorbent by using a graduated cylinder and weighed the absorbent before use. The absorbent was then placed into the upper canister, and glass balls were placed into the lower canister as filler (10). Except for the canister size, the protocols were the same as Study 1. Six trials were performed for each absorbent.
During anesthesia, inspired and end-tidal CO2 concentrations, inspiratory and expiratory minute volume (MVins and MVexp, respectively), and V̇Eco2 were monitored with a CS/3. These data were recorded at 1-min intervals. The CO2 absorption capacity of each absorbent was evaluated by calculating the sum of low-flow anesthesia time and the volume of CO2 exhaled by the patient (CO2 absorption of capacity [L/100 g absorbent]) until CO2 rebreathing occurred. Because the fraction of V̇Eco2 that is recirculated depends on the minute volume, the volume of CO2 recirculated (V̇Rco2) was calculated as follows:MATH
Values are expressed as the mean ± sd. Intergroup comparisons of data were performed with one-way analysis of variance and Tukey’s post hoc test. A P value of <0.05 was considered to be statistically significant.
Thirty-four patients were enrolled in Study 1. Nine of these patients belonged to at least two of the absorbent groups. In Study 2, 35 patients were enrolled, and 12 patients belonged to at least two of the absorbent groups. After the first use, the color of Amsorb remained light violet-blue, whereas Medisorb and Sodasorb regained their white color. Table 1 indicates the number of patients for which the same absorbent was used, the number of times, and the V̇Rco2 until CO2 rebreathing occurred in the three groups in both Study 1 and Study 2. In Study 1, the longevity of Amsorb was 213 ± 71 min, significantly less than those of Medisorb (445 ± 125;P < 0.01) and Sodasorb (503 ± 89;P < 0.001) (Table 1). The CO2 absorption capacity (L/100 g absorbent) of Amsorb was 5.5 ± 1.2, significantly less than those of Medisorb (10.7 ± 1.7) and Sodasorb (12.1 ± 1.8;P < 0.001) (Fig. 1). In Study 2, the longevity of Amsorb was 218 ± 61 min, significantly less than those of Medisorb (533 ± 136) and Sodasorb (528 ± 103;P < 0.001) (Table 1). The CO2 absorption capacity (L/100 g absorbent) of Amsorb was 7.6 ± 1.6, significantly less than those of Medisorb (14.4 ± 1.8) and Sodasorb (14.8 ± 2.3;P < 0.001) (Fig. 1). There was a significant difference in the CO2 absorption capacity (L/100 g absorbent) of Medisorb between Study 1 and Study 2 (P < 0.05). There was, however, no significant difference in the CO2 absorption capacity of Amsorb or Sodasorb between Study 1 and Study 2.
In this study, the time to exhaustion and the volume of CO2 absorption of Amsorb were 50% less than those of soda lime with either the small or large canisters. The extent to which the CO2 absorption capacity of Amsorb was less than that of soda lime was comparable with in vitro studies (5–8), except for the study by Murray et al. (1). The experimental conditions used in this study were chosen to reflect routine clinical anesthesia. We filled the canister with a CO2 absorbent, allowing the absorbent to settle first. If the CO2 absorbent did not become exhausted, the same CO2 absorbent was used again to absorb CO2 after a period of rest. This was continued until the CO2 absorbent became completely exhausted. The CO2 absorption capacity was determined without a period of rest until the inspired CO2 concentration reached 2 mm of mercury. During rest, however, alkali hydroxides are regenerated. After a 24-hour rest, the concentration of NaOH returned to almost the preexposed value (7). Therefore, the color change is reversible and is not a good indicator of the exhaustion of soda lime, whereas with Amsorb, color change is not reversible and does indicate exhaustion because Amsorb does not contain a strong alkali. Jorgensen and Jorgensen in 1977 (11) reported that after a 24-hour rest, exhausted CO2 absorbent (600 g) absorbed CO2 again for approximately 40 minutes and that this regeneration was of no clinical importance. The findings of this study are consistent with their report, suggesting that the regeneration of the alkali hydroxides is not of much importance for the CO2 consumption capacity of absorbents because the CO2 absorption capacity between Amsorb and soda lime in the in vitro and in vivo studies was not significantly different.
Bedi et al. (8) recently reported that the CO2 absorption capacity of CO2 absorbents differed after changing the absorbent canister. They compared the CO2 absorption capacity of Amsorb with that of Medisorb by using four CO2 absorbent canisters in vitro: an ADU compact canister, a ThermH2O canister (Raincoat Corporation, Louisville, KY), a Dräger Julian Anesthetic machine/canister (Dräger, Luebeck, Germany), and an Ohmeda Modulus Anesthetic machine/canister (Datex-Ohmeda, Helsinki, Finland). They reported that CO2 absorption capacities of Amsorb and Medisorb with the Ohmeda Modulus Anesthetic machine/canister were six times higher than those with the ADU compact canister. They also stated that the difference in the CO2 absorption capacity between Amsorb and Medisorb was small compared with the differences caused by the type of canister. Our findings that the difference in the CO2 absorption capacity between Amsorb and Medisorb was almost constant are consistent with those in the study by Bedi et al. (8). CO2 absorption capacities (L/100 g of absorbent) of Amsorb and Medisorb with the Ohmeda Aestiva 3000 Anesthetic machine/canister, which has the same volume and design as the Ohmeda Modulus Anesthetic machine/canister, were only 25% higher than those with an ADU compact canister in this study. The reason for the lack of canister-dependent differences of canister on the CO2 absorption capacity in this study is unclear. These differences might be accounted for by the difference in the volume of absorbents to fill the canister (12). We used 700 mL (approximately 550 g) of CO2 absorbent in the Ohmeda canister, whereas Bedi et al. (8) used 1000 g of CO2 absorbent.
Although the results of this clinical study qualitatively confirm those of the in vitro studies (1,5–8), which reported that the longevity and CO2 absorption capacity of Amsorb were less than those of soda lime, there is a quantitative disagreement in CO2 absorption of Amsorb and soda lime between the previous in vitro studies and our in vivo and in vitro studies. Table 2 provides comparisons of the results reported in this study and the previous in vitro studies (1,5,7,8), which are limited to the results of Amsorb and Medisorb used in an ADU compact canister and a large canister. For CO2 absorption of Amsorb with a large canister, Murray et al. (1) reported that 100 g of Amsorb absorbed 11 L of CO2; Stabernack et al. (5) reported approximately 13 L absorbed; Ueyama et al. (7) reported 18 L; Bedi et al. (8) reported 16 L; and in this study we report only 7 L. However, for Medisorb with an Ohmeda canister, the CO2 absorption capacity (L/100 g of absorbent) of Medisorb in the study by Ueyama et al. (7), who used an Ohmeda Excel canister (Datex-Ohmeda, Madison, WI), and Bedi et al. (8), who used an Ohmeda Modulus canister (Datex-Ohmeda), was 35 and 23 L, respectively, whereas the corresponding value in this study was 15 L. Possible explanations for these discrepancies might be the different end points used to define breakthrough. In this study, rebreathing during general anesthesia was defined as an increased inspired pressure of CO2 of more than 2 mm of mercury (9), whereas the others defined it at 5 mm of mercury (1,5–7) or 0.5%(8). Assuming that the ability to absorb CO2 is a linear function for both Amsorb and Medisorb, the CO2 absorption capacity of Amsorb and Medisorb until a 5 mm of mercury end point in our study with a large canister would be 17 L and 38 L, respectively—values that are comparable with those in the studies by Stabernack et al. (5), Ueyama et al. (7), and Bedi et al. (8). There is no evidence, however, that the ability to absorb CO2 is a linear function for all of the absorbents. It is unclear why there is less absorption with Amsorb and Medisorb with an ADU compact canister in the study by Bedi et al. (8) than in this study. In addition, there are no obvious explanations for the lower absorption of Amsorb and the smaller difference between the CO2 absorption capacity of Amsorb and that of soda lime reported by Murray et al. (1) than reported by others (5,7,8). The difference between continuous flow and the ventilation might be associated with the lower absorption and difference reported by Murray et al. (1). Murray et al. (1) used continuous flow, whereas the lung was ventilated in other studies (5,7,8) and in our study. It is possible that the pause during breathing allows for more complete CO2 removal.
The cost issue of Amsorb must be considered because it is expensive (4,6). If absorbents are used until exhaustion over a period of one week, our data suggest that the cost of using Amsorb under low-flow anesthesia might be exaggerated because the longevity of Amsorb is half that of soda lime. Baum and Van Aken (6) stated, however, that if the absorbent were routinely changed once a week, unless it had been exhausted, the cost of using Amsorb might not be exaggerated. In most practices, CO2 absorbent is replaced according to a routine schedule (12). However, Baum and Van Aken (6) reported that on the basis of their clinical experience, the scavenging capacity of 1.5 L of Amsorb contained in a jumbo canister was sufficient to absorb CO2 safely over a period of an entire week. Thus, the cost issue of Amsorb is not straightforward because it depends on the purchase cost and the schedule for changing CO2 absorbents.
Several limitations in this study must be addressed. First, in Study 2, only 700 mL of absorbent was used in the upper canister, and the lower canister was not filled with absorbent because we wished to complete the trial within a week. As mentioned above, the absorbent’s packing might greatly affect the CO2 absorption capacity of each absorbent. Therefore, our result with a large canister might not reflect normal clinical practice. Second, it is possible that the results obtained in this study, in which inhaled anesthetics were excluded from the anesthetic protocol, might not apply to clinical conditions, in which volatile anesthetics, especially sevoflurane, are used, because we did not use volatile anesthetics. Amsorb is chemically unreactive with volatile anesthetics. However, soda lime degrades sevoflurane to Compound A and desflurane, enflurane, and isoflurane to CO (1,3,4,10). Knolle and Gilly (13) reported that the presence of 0.5% isoflurane significantly reduced the CO2 absorption capacity of dried soda lime, compared with the absence of isoflurane, and concluded that CO formation and CO2 absorption compete for the same reactant. It is unlikely, however, that the degradation of sevoflurane to Compound A would decrease the longevity of standard (moist) soda lime, because degradation with dry absorbent is enormously more than degradation with standard absorbent (14).
In summary, the CO2 absorption capacity of Amsorb is half that of soda lime under clinical low-flow (1 L/min) anesthesia with either a small or large canister. Further study with a larger amount of absorbent, which is used in clinical practice, is required regarding the longevity of Amsorb and soda lime.
1. Murray JM, Renfrew CW, Bedi A, et al. Amsorb: a new carbon dioxide absorbent for use in anesthetic breathing systems. Anesthesiology 1999; 91: 1342–8.
2. Kharasch ED. Putting the brakes on anesthetic breakdown [editorial]. Anesthesiology 1999; 91: 1192–4.
3. Yamakage M, Yamada S, Chen X, et al. Carbon dioxide absorbents containing potassium hydroxide produce much larger concentrations of compound A from sevoflurane in clinical practice. Anesth Analg 2000; 91: 220–4.
4. Higuchi H, Adachi Y, Arimura S, et al. Compound A concentrations during low-flow sevoflurane anesthesia correlate directly with the concentration of monovalent bases in carbon dioxide absorbents. Anesth Analg 2000; 91: 434–9.
5. Stabernack CR, Brown R, Laster MJ, et al. Absorbents differ enormously in their capacity to produce compound A and carbon monoxide. Anesth Analg 2000; 90: 1428–35.
6. Baum J, Van Aken H. Calcium hydroxide lime: a new carbon dioxide absorbent—a rationale for judicious use of different absorbents [editorial]. Eur J Anaesthesiol 2000; 17: 597–600.
7. Ueyama H, Takashima M, Suzuki T, et al. Warning: carbon dioxide absorption capacity of Amsorb was unexpectedly low in low-flow anesthesia [letter]. Anesthesiology 2000; 93: 1560–1.
8. Bedi A, Gallagher AC, Murray JM, Fee H. The in vitro performance of carbon dioxide absorbents with and without strong alkali [abstract]. Anesthesiology 2000; 93: A582.
9. van Genderingen HR, Gravenstein N, van der Aa JJ, Gravenstein JS. Computer-assisted capnogram analysis. J Clin Monit 1987; 3: 194–200.
10. Bito H, Ikeuchi Y, Ikeda K. Effects of the water content of soda lime on compound A concentration in the anesthesia circuit in sevoflurane anesthesia. Anesthesiology 1998; 88: 66–71.
11. Jorgensen B, Jorgensen S. The 600 gram CO2
absorption canister: an experimental study. Acta Anaesthesiol Scand 1977; 21: 637–44.
12. Ohrn M, Gravenstein N, Good ML. Duration of carbon dioxide absorption by soda lime at low rates of fresh gas flow. J Clin Anesth 1991; 3: 104–7.
13. Knolle E, Gilly H. Absorption of carbon dioxide by dry soda lime decreases carbon monoxide formation from isoflurane degradation. Anesth Analg 2000; 91: 446–51.
© 2001 International Anesthesia Research Society
14. Funk W, Gruber M, Wild K, Hobbhahn J. Dry soda lime markedly degrades sevoflurane during simulated inhalation induction. Br J Anaesth 1999; 82: 193–8.