Psychomotor and Clinical Assessment of Flumazenil as an Antagonist of Midazolam

The benzodiazepine antagonist flumazenil (0.01 mg/kg) has been compared with doxapram (1 mg/kg) and saline for the reversal of anaesthesia with intravenous midazolam, alfentanil, nitrous oxide in oxygen and isoflurane. The completeness of reversal was assessed by means of a four-choice reaction time test, 1 and 3 h following the antagonist. In addition, the level of sedation was graded using a five-point scale.

Psychomotor testing showed that 60 min after administration of the antagonist, there were marked increases in reaction times (P < 0.05) both in the control and doxapram groups, but not in those receiving flumazenil. At 180 min, however, reaction times in all groups had returned to baseline values. In contrast, there was a significant difference in the sedation scores between the saline and flumazenil groups throughout the study period (P < 0.05).

During the 4 h following midazolam, there was no evidence of re-sedation in any of the groups despite the relatively high midazolam dosage.

Prospective Study of Liver Function following Repeat Halothane and Enflurane

In a prospective study of liver function following repeat anaesthesia, patients who received repeat halothane had a higher frequency of abnormal liver enzyme results than a similar group who received repeat enflurane. Obesity and short intervals between administrations increased the likelihood of abnormal liver enzyme activity in the halothane group. Enflurane would seem to be the volatile agent of choice for repeat anaesthesia in such circumstances.

Intrathecal ropivacaine or bupivacaine with fentanyl for labour

Combined spinal-epidural (CSE) is widely used to provide pain relief in labour while minimizing motor blockade. Aiming to further reduce associated motor weakness, we compared ropivacaine 2.5 mg in the intrathecal injection with a standard bupivacaine CSE in a double-blind study. Forty women were randomized to receive either bupivacaine 2.5 mg or ropivacaine 2.5 mg intrathecally, both with fentanyl 0.025 mg. There were no significant differences between the groups regarding the onset, duration or quality of analgesia or the level of sensory block attained. Forty per cent of the women (8/20) receiving bupivacaine developed detectable motor block compared with only 5% (1/20) in the ropivacaine group (P<0.05). Vibration sense was impaired in one woman in each group. Adverse effects did not differ between groups. We conclude that intrathecal ropivacaine 2.5 mg in combination with fentanyl 0.025 mg as part of a CSE technique provides rapid and safe analgesia for labour as effective as that achieved with bupivacaine 2.5 mg and with significantly less motor block.

Patient-controlled analgesia for labour using remifentanil: a feasibility study

We have investigated the efficacy and safety of remifentanil in a patient-controlled analgesia device for labour in 21 women. Remifentanil was available in increasing doses (bolus doses 0.25-1.0 microg x kg(-1)) with and without a background infusion (0.025-0.05 microg x kg(-1) x min(-1)). A lockout time of 2 min was used. Thirteen out of 21 (62%) women chose to continue using remifentanil up to and during delivery. Nineteen out of 21 (90%) achieved a reduction in pain score from baseline. Using a VAS of 0-10 cm the median maximum reduction in pain score was 3 cm (range 0-8 cm). There was a significant reduction (P<0.05) from baseline pain scores (median= 8 cm) to scores at bolus doses in the range 0.25-0.5 microg x kg(-1) (median=5 cm). There were no significant reductions in the fetal heart rate. Apgar scores and cord blood gas analyses remained within normal limits. We conclude that a remifentanil patient-controlled analgesia system (bolus doses 0.25-0.5 microg x kg(-1), without a background infusion) may safely provide worthwhile, although incomplete, analgesia for labour.

The in vitro performance of carbon dioxide absorbents with and without strong alkali apparatus

We report the in vitro longevity of a conventional soda lime carbon dioxide absorbent and an absorbent free from strong alkali (Amsorb). Although the times taken to breakthrough of carbon dioxide (> 0.5%) within an in vitro low flow breathing system were shorter with the alkali-free absorbent, we found that the size and shape of the absorbent container was the major factor in determining the efficiency of the carbon dioxide absorbents.

A qualitative investigation into the physical stability of polypropylene and polyethylene in liquid isoflurane and sevoflurane

The interaction between medical plastics and drugs is complex. Drug absorption into plastics may affect drug dosage and the migration of plastics' additives into a drug solution may affect drug composition. We investigated the stability of those plastics which may be used in infusion systems to inject liquid volatile anaesthetic drugs directly into an anaesthetic breathing system. Samples of two types of polypropylene from a syringe barrel and plunger and low- and high-density polyethylene from extension tubing were exposed to isoflurane and sevoflurane for 1, 7 or 250 days. All samples were from the same batches. Samples of the plastics (n = 24) and the liquid volatile anaesthetics (n = 24) were subjected to Fourier transform-infrared spectroscopy to produce series of absorption spectra. By reference to control sample absorption spectra, this allows detection of anaesthetic drug absorption into the plastics or migration of the plastics or their additives into the liquid anaesthetics. We found no evidence of migration of the plastic components or their additives into the liquid anaesthetic drugs at any of the exposure periods. Similarly, we found no evidence of absorption of isoflurane or sevoflurane by any of the plastic components during short-term exposure of either 1 or 7 days. However, there was evidence of some absorption of the anaesthetic drugs by the polyethylene plastics after about 8 months' exposure. It would appear that low- and high-density polyethylene and polypropylene are suitably safe for use in infusion systems for the direct injection of isoflurane and sevoflurane into anaesthetic breathing systems.

Tracheal intubating conditions after induction with sevoflurane 8% in children. A comparison with two intravenous techniques

We studied tracheal intubating conditions in 120 healthy children, aged 3-12 years, in a blinded, randomised clinical trial. Children were randomly allocated to one of three groups: group PS, propofol 3 mg.kg-1 and succinylcholine 1 mg.kg-1 (n = 40); group PA, propofol 3 mg.kg-1 and alfentanil 10 microg.kg-1 (n = 40); group SF, sevoflurane 8% in 60% nitrous oxide in oxygen for 3 min (n = 40). Tracheal intubating conditions were graded according to ease of laryngoscopy, position of vocal cords, coughing, jaw relaxation and movement of limbs. Overall intubating conditions were acceptable in 39 of 40 children in the propofol/succinylcholine group, 21 of 40 children in the propofol/alfentanil group and 35 of 40 children in the sevoflurane group. Children receiving propofol and succinylcholine or sevoflurane had better intubating conditions overall than those given propofol and alfentanil (p < 0.01). In conclusion, anaesthetic induction and tracheal intubation using sevoflurane 8% for 3 min is a satisfactory alternative to propofol with succinylcholine in children.

Amsorb: a new carbon dioxide absorbent for use in anesthetic breathing systems

BACKGROUND

This article describes a carbon dioxide absorbent for use in anesthesia. The absorbent consists of calcium hydroxide with a compatible humectant, namely, calcium chloride. The absorbent mixture does not contain sodium or potassium hydroxide but includes two setting agents (calcium sulphate and polyvinylpyrrolidine) to improve hardness and porosity.

METHODS

The resultant mixture was formulated and subjected to standardized tests for hardness, porosity, and carbon dioxide absorption. Additionally, the new absorbent was exposed in vitro to sevoflurane, desflurane, isoflurane, and enflurane to determine whether these anesthetics were degraded to either compound A or carbon monoxide. The performance data and inertness of the absorbent were compared with two currently available brands of soda lime: Intersorb (Intersurgical Ltd., Berkshire, United Kingdom) and Dragersorb (Drager, Lubeck, Germany).

RESULTS

The new carbon dioxide absorbent conformed to United States Pharmacopeia specifications in terms of carbon dioxide absorption, granule hardness, and porosity. When the new material was exposed to sevoflurane (2%) in oxygen at a flow rate of 1 l/min, concentrations of compound A did not increase above those found in the parent drug (1.3-3.3 ppm). In the same experiment, mean +/-SD concentrations of compound A (32.5 +/- 4.5 ppm) were observed when both traditional brands of soda lime were used. After dehydration of the traditional soda limes, immediate exposure to desflurane (60%), enflurane (2%), and isoflurane (2%) produced concentrations of carbon monoxide of 600.0 +/- 10.0 ppm, 580.0 +/- 9.8 ppm, and 620.0 +/-10.1 ppm, respectively. In contrast, concentrations of carbon monoxide were negligible (1-3 ppm) when the anhydrous new absorbent was exposed to the same anesthetics.

CONCLUSIONS

The new material is an effective carbon dioxide absorbent and is chemically unreactive with sevoflurane, enflurane, isoflurane, and desflurane.

Ventilatory characteristics in mechanically ventilated patients during manual hyperventilation for chest physiotherapy

We measured the tidal volumes and peak inflation pressures generated during manual hyperventilation for chest physiotherapy in 25 adult ventilated patients. The average tidal volume ranged from 711 to 1511 ml, with a mean (SD) of 1120 (274) ml. There was a negative correlation (p < 0.05) between the average tidal volume and the lung injury, as measured by the Murray lung injury score. The average peak inflation pressure ranged from 37 to 74 cmH2O with a mean (SD) of 51.5 (7.6) cmH2O. There was a positive correlation (p < 0.05) between average peak inflation pressure and the lung injury score. Mean (SD) PaO2 improved by 18.3 (14.3) kPa from baseline after physiotherapy (p < 0.05). Mean (SD) PaCO2 decreased by 0.1 (0.4) kPa. As the lung score increases higher inflation pressures and smaller tidal volumes are used suggesting an increased potential for barotrauma or volutrauma in susceptible lungs.

An in vitro evaluation of flow from multihole epidural catheters during continuous infusion with four different infusion pumps

We have observed in vitro the distribution of flow from 10 identical multihole epidural catheters during continuous infusion with four different infusion pumps. The pumps chosen were the B Braun Perfusor Secura FT syringe driver and three volumetric infusion pumps utilising different pumping mechanisms (Dekra 3000 BL, Graseby 500 and CADD-Prizm). These pumps infused 0.9% saline through each catheter at 5 ml.h-1, 15 ml.h-1, 50 ml.h-1 and 99 ml.h-1 for 3 min. The number of holes through which flow occurred and the catheter hole where flow predominated during each test were recorded. The pressure waveform generated during each infusion was displayed and the peak pressure recorded. In 38 of the 160 tests (24%) the largest proportion of flow was seen at the hole closest to the catheter tip. The CADD pump generated multihole flow during significantly more tests (p < 0.0001) than the other pumps and produced significantly higher driving pressures (p < 0.001) at all infusion rates compared with the Graseby and Perfusor pumps. The CADD was the only pump to produce flow from all three holes of the catheter at 5 ml.h-1.

Balanced pre-emptive analgesia: does it work? A double-blind, controlled study in bilaterally symmetrical oral surgery

We studied 32 patients undergoing bilateral symmetrical lower third molar surgery under general anaesthesia to determine if the combined effects of pre-emptive local anaesthetic block using 0.5% bupivacaine, together with i.v. tenoxicam and alfentanil had any benefits over postoperative administration. Patients acted as their own controls and were allocated randomly to have surgery start on one side, the second side always being the pre-emptive side. Difference in pain intensity between the two sides was determined using visual analogue scales completed by each individual at 6 h, and at 1, 3 and 6 days after operation. A long-form McGill pain questionnaire was also used to assess difference in pain intensity between the two sides on the morning after surgery. There was no significant difference in pain intensity at any time after surgery. Our findings indicate that the combined use of pre-emptive analgesia from 0.5% bupivacaine, tenoxicam and alfentanil did not reduce postoperative pain intensity in patients undergoing molar exodontia.

Double-blind comparison of extradural block with three bupivacaine-ketamine mixtures in knee arthroplasty

We have compared 0.5% bupivacaine 75 mg (group A; n = 15) with three 0.5% bupivacaine 75 mg-ketamine mixtures for extradural block in 59 ASA I-III patients undergoing total knee replacement in a randomized, double-blind study. The following doses of preservative-free 1% ketamine were used: 0.3 mg kg-1 (group B: n = 14); 0.5 mg kg-1 (group C: n = 5); and 0.67 mg kg-1 (group D: n = 15). Level of sensory block, degree of motor weakness and sedation scores were recorded before and after operation. Duration of postoperative analgesia was also noted. There was no difference between groups in median maximum level of sensory block (group A: T4 (range T10-T2); group B: T4 (T10-T2); group C: T4 (T8-T2); and group D: T3 (T8-C3)) or in the degree of motor block. Thirty-three of the 44 patients who received ketamine showed signs of systemic absorption (blurred vision, sedation) within 10 min of injection. There was no significant difference between groups in median duration of analgesia (group A: 240 (range 115-340) min; group B: 198 (97-460) min; group C: 150 (122-448) min; and group D: 210 (130-390) min). No patient suffered any adverse psychomimetic effects. We conclude that at the doses used, addition of ketamine to extradural bupivacaine did not improve extradural block in adult patients undergoing total knee replacement.

Cardiovascular effects of 50% nitrous oxide in older adult patients anaesthetized with isoflurane or halothane

We have studied the cardiovascular effects of equipotent concentrations of halothane or isoflurane, with or without 50% nitrous oxide in oxygen, in 80 patients, aged 60 yr or more, during 20 min of stable equipotent anaesthesia. Non-invasive measurement techniques were used, with suprasternal Doppler ultrasonography for estimating cardiac output. Both isoflurane and halothane reduced heart rate and systemic arterial pressure. With isoflurane, mean rate decreased from 72 (SD 9.7) to 67 (10.4) beat min-1 and with halothane from 76 (10.1) to 65 (9.1) beat min-1 (P < 0.05). Mean diastolic arterial pressure decreased from 81 (11.3) to 58 (17.0) mm Hg with isoflurane and from 86 (14.7) to 70 (13.3) mm Hg with halothane (P < 0.05). Cardiac index decreased from 3.1 (1.03) to 2.7 (0.71) litre min-1 m-2 with isoflurane and from 3.1 (0.98) to 2.5 (0.57) litre min-1 m-2 with halothane (P < 0.05). Systemic vascular resistance decreased significantly in all groups except those receiving halothane with nitrous oxide. Nitrous oxide resulted in significantly less depression of cardiac index when given with isoflurane than when given with halothane. The mean percentage change in cardiac index during isoflurane anaesthesia without nitrous oxide was 16.7%; with nitrous oxide there was a 0.5% increase. Halothane, in combination with nitrous oxide, resulted in greater depression of cardiac index than isoflurane with nitrous oxide. The mean percentage change with halothane was 20.4% (22.2%); with isoflurane there was a 0.5% (27.1%) increase (P < 0.05). Hypotension was more pronounced in patients anaesthetized with isoflurane (n = 40) than those anaesthetized with halothane (n = 40), irrespective of the presence of nitrous oxide. The mean percentage decrease with isoflurane was 29.7% (21.10%) compared with 16.8% (16.78%) with halothane (P < 0.05).

Comparative tolerability profiles of the inhaled anaesthetics

Improved understanding of the structure/activity relationship of inhaled anaesthetics has resulted in the synthesis of fluorinated compounds which are more potent and less toxic than their unfluorinated antecedents. The toxic effects of inhaled anaesthetics on the liver and kidney are complex but, in general, are related to the extent to which individual inhaled agents are metabolised. Halothane hepatotoxicity is a rare, idiosyncratic reaction which typically occurs in obese women having more than one exposure to the drug within a short time interval. All currently available volatile anaesthetic drugs have depressant effects on the cardiovascular and respiratory systems; arrhythmias are more likely with halothane than with the fluorinated ethers. Cerebral blood flow tends to increase during inhalation anaesthesia, especially with halothane and in the presence of hypercarbia; isoflurane may be given sparingly during neurosurgical procedures whilst monitoring its end-tidal concentration. Although the volatile agents tend to cause uterine relaxation they may be given safely in low concentration to avoid awareness during Caesarean section. In general, young children require rather higher concentrations of volatile agents than adults and seem to be less susceptible to organ toxicity. Two relatively new volatile agents, sevoflurane and desflurane, offer some advantages over isoflurane but neither is an "ideal drug'. Sevoflurane interacts with soda-lime to produce a series of degradation products, the most important of which is compound A. Production is greatest during low-flow, closed circuit anaesthesia using high inspired concentrations of the drug. Compound A has nephrotoxic potential in rats but the clinical significance of the interaction between sevoflurane and soda-lime is unclear. Nitrous oxide when given for prolonged periods may cause irreversible bone marrow depression.

Serum glutathione S-transferase concentrations and creatinine clearance after sevoflurane anaesthesia

The effects of sevoflurane and isoflurane on serum glutathione S-transferase concentrations and creatinine clearance were compared in 50 ASA I-III patients aged over 18 years undergoing body surface surgery of 1-3 h predicted duration. Patients randomly received sevoflurane (n = 24) or isoflurane (n = 26) in nitrous oxide and oxygen (FIO2 = 0.4) via a nonrebreathing system. Fluids were standardised and patient's lungs ventilated to normocapnia. Expired concentration of anaesthetic agent was adjusted to maintain systolic arterial pressure between 70 and 100% of baseline. Patients received significantly less (p < 0.05) sevoflurane (1.0 MAC-h) than isoflurane (1.5 MAC-h). Using serum glutathione S-transferase concentrations and creatinine clearance as markers of hepatic and renal function respectively, no statistically significant differences were identified between the groups.

A comparison of intrathecal morphine-6-glucuronide and intrathecal morphine sulfate as analgesics for total hip replacement

Postoperative analgesia was assessed after intrathecal administration of morphine-6-glucuronide (M6G) 100 micrograms and 125 micrograms in 75 patients undergoing total hip replacement. Analgesia was excellent and was similar to that obtained after intrathecal administration of morphine sulfate 500 micrograms. Visual analog pain scores recorded postoperatively were low (median = 0) and were similar in all three groups. However, at 6 and 10 h after operation significantly more patients in the M6G 125 group recorded pain as 0 compared with patients in the morphine group (P < 0.04, P < 0.01) and significantly more patients in the M6G 100 group recorded pain as 0 at 24 h after operation compared with patients in the morphine group (P < 0.04). Postoperative meperidine consumption using a patient-controlled system was also similar in each of the three treatment groups. Nausea and emesis occurred frequently in all groups; morphine (nausea 88%, vomiting 76%), M6G 100 micrograms (nausea 76%, vomiting 64%), and M6G 125 micrograms (nausea 88%, vomiting 60%). Respiratory depression occurred in two and three patients, respectively, in the M6G 100-microgram and 125-microgram groups but did not occur in any patient who received morphine sulfate. The lack of statistical significance in the difference in incidence of respiratory depression between the groups may represent a type II error. However, the risk of late respiratory depression developing after administration of any intrathecal opioid necessitates careful postoperative observation of patients. As M6G is a potent intrathecal analgesic further investigation of this drug using small doses may be useful.

An investigation of the potential morphine sparing effect of midazolam

The effect of a bolus and continuous infusion of midazolam on postoperative morphine consumption was assessed in a placebo-controlled, double-blind, randomly allocated trial of 50 patients undergoing elective abdominal hysterectomy. Patients in the trial group received a bolus dose of midazolam 5 mg.70 kg-1 at induction followed by an infusion at a rate of 1 mg.70 kg-1.h-1 over the next 48 h. Morphine consumption in the midazolam group was significantly lower in the first 12 h postoperatively (p < 0.02) but there was no significant difference between the two groups thereafter. Patients in the midazolam treated group had lower pain scores over the first 24 h. Also, a significantly greater number of patients in the midazolam group required no antiemetic medication over the 48 h study period (p < 0.05). Assessment of sedation revealed no significant difference between groups. We conclude that low dose midazolam has a significant, but short-lived, morphine sparing effect.

Effect of the addition of alfentanil to lignocaine during axillary brachial plexus anaesthesia

Peripheral administration of opioids has been suggested as a means of improving regional block. We studied 60 patients receiving axillary brachial plexus anaesthesia, allocated randomly to receive either normal saline 10 ml or normal saline 10 ml with alfentanil 10 micrograms/kg body weight through an axillary cannula. All patients received 1.5% lignocaine at a dose of 7 mg/kg body weight with adrenaline 1 in 200,000. The incidence of satisfactory block was similar in both groups. Although the percentage of patients with complete anaesthesia in the median nerve distribution was greater in the alfentanil group, there was no significant difference in any other distribution. The time to return of sensation and motor function was prolonged significantly in the alfentanil group (P < 0.05). After return of normal sensation, there was no significant difference between groups in postoperative analgesia. In a second part of the study, there was no significant increase in plasma concentrations of alfentanil in 10 patients given lignocaine and alfentanil, as outlined above. These observations suggest that alfentanil may have a peripheral local anaesthetic action.

The provision of adult intensive care in Northern Ireland with reference to the role of high dependency care

In 1991 an audit of Intensive Care Services was carried out by the Northern Ireland Intensive Care Group. In conjunction with this regional overview, all patients in the Regional Intensive Care Unit, (RICU) in the Royal Victoria Hospital were assessed daily, over a 10 month period in 1990-91 and classified as conforming to either intensive care or high dependency status. These data were then used to compare adult intensive care service in Northern Ireland with recent national and international recommendations on intensive care. Ten units in Northern Ireland were surveyed. In regard to national or international guidelines, all ten were deficient to some degree. Four units had significant deficiencies; small patient numbers, lack of 'dedicated' 24 hr medical cover and or deficiencies in the provision of appropriate monitoring and or equipment. There was a large diversity in casemix among the ten units surveyed which suggested differing admission criteria. The bed occupancy of RICU was 100%. Refused admissions constituted a further 13% of unresourced workload. The lack of physically separate, dedicated high dependency unit facilities meant that 26% of bed days were devoted to HDU care (usually for "improved" intensive care unit patients not yet ready for discharge to a general ward. Achieving nationally recommended intensive care standards (on a regional basis) is probably only possible if a number of the smaller intensive care units are redesignated as high dependency units, and patients requiring intensive care are concentrated in a smaller number of larger ICUs. This will increase the frequency of interhospital transfer of critically ill patients.

The effect of alkalinisation of lignocaine on axillary brachial plexus anaesthesia

Alkalinisation of local anaesthetic drugs is a controversial technique for improving regional blockade. Forty-two patients scheduled for upper limb surgery received axillary brachial plexus anaesthesia using a cannula technique. Patients were randomly allocated to receive either lignocaine 1.5% with 1 in 200,000 adrenaline (pH = 4.2) or lignocaine 1.5% with 1 in 200,000 adrenaline (pH = 7.2). There was no significant difference in the incidence of satisfactory blockade or distribution of anaesthesia between the two groups. The percentage of patients with complete anaesthesia at 10, 20 and 30 min following injection was significantly increased in the alkalinized group with regard to the ulnar and median nerves, and the median cutaneous nerve of the arm (p < 0.05). In the alkalinized group, there was a significant reduction in the time to useful anaesthesia and a reduced requirement for adjuvants (p < 0.05). There was no effect on the duration of anaesthesia.

Anaesthesia and adverse effects after intrathecal pethidine hydrochloride for urological surgery

Anaesthesia, postoperative analgesia and the incidence of adverse effects after intrathecal pethidine hydrochloride 0.50 mg.kg-1 and 0.75 mg.kg-1 were assessed and compared with a conventional technique using isobaric bupivacaine 13.75 mg in patients undergoing transurethral resection of the prostate gland. Sensory and motor block were significantly shorter with both pethidine regimens than with bupivacaine (p < 0.001). Although sensory and motor block were shorter after pethidine 0.50 mg.kg-1 than after pethidine 0.75 mg.kg-1 the difference in duration was clinically insignificant. The incidence of incomplete motor block was significantly greater with pethidine 0.50 mg.kg-1 than with bupivacaine 13.75 mg.kg-1 (p < 0.01). Visual analogue pain scores recorded after the operation were low and were similar in the pethidine groups to those obtained with bupivacaine alone. Mean arterial blood pressure was significantly lower in both pethidine groups compared with the bupivacaine group between 30 and 240 min after intrathecal injection (p < 0.001). However, the within-group reductions in blood pressure were within clinically acceptable limits. The incidences of nausea and emesis were low and emesis occurred in patients in the bupivacaine group only (p < 0.03). Pruritus was seen only in patients receiving pethidine. Intra-operative sedation occurred more often in patients receiving both pethidine 0.50 mg.kg-1 and 0.75 mg.kg-1 compared with patients receiving bupivacaine (p < 0.04). Both pethidine regimens provided acceptable anaesthesia and there were no significant differences between the two regimens in quality of intra-operative anaesthesia, incidence of adverse events or postoperative analgesia.

Molecular sieves: an alternative method of carbon dioxide removal which does not generate compound A during simulated low-flow sevoflurane anaesthesia

Molecular sieves are used in industry to 'scrub' industrial gases. We examined, during simulated low-flow closed system anaesthesia, (1) the carbon dioxide adsorbing potential of molecular sieves and (2) the reactivity of the sieves compared to soda lime using sevoflurane as an indicator. A low-flow anaesthetic system containing 13X molecular sieves was connected to a model lung. End-tidal concentrations of CO2 were measured continuously at an O2 flow of 800 ml.min-1 and a CO2 flow of 200 ml.min-1. In the second study, sevoflurane (FE'sevo 1.7%) was added to the system after which samples were taken from the inspiratory limb of the anaesthetic system. This experiment was performed both during carbon dioxide removal with soda lime and with the molecular sieves. The samples were stored in gas-tight syringes and analysed by gas chromatography. The temperature of both absorbents was measured throughout the study. The molecular sieves adsorbed carbon dioxide (20%) efficiently for a period of 5 h. There was a gradual increase from the baseline of 4.4% to 4.5, 5.4, and 6.0% at 90, 180, and 300 min, respectively. When sevoflurane was added to the system, compound A was detected at the start of both experiments. However, when soda lime was used the concentrations of compound A increased 10-fold after 2.5 h compared with baseline values. No increase in compound A was observed when molecular sieves were used for carbon dioxide removal. The highest mean (SD) temperature of the molecular sieves was 41.5 (3.2) degrees C. Molecular sieves are effective adsorbents of carbon dioxide when used in a simulated low-flow, closed anaesthetic system.(ABSTRACT TRUNCATED AT 250 WORDS)