Intravenous boluses of fentanyl, 1 µg kg⁻¹, and remifentanil, 0.5 µg kg⁻¹, give similar maximum ventilatory depression in awake volunteers.

BACKGROUND: The relative respiratory effects of fentanyl and remifentanil, administered as i.v. bolus, have not previously been studied. We determined what remifentanil bolus dose gave the same maximum depression of ventilation as 1 µg kg(-1) of fentanyl. METHODS: Twelve healthy volunteers rebreathed in a system designed to dampen variations in end-tidal carbon dioxide tension PE'CO2 so that measurements would be obtained at similar levels of CO(2) stimulation. The minute ventilation was measured before (V(preinj)) and after injection (V(nadir)) of fentanyl, 1 µg kg(-1), and remifentanil, 0.25, 0.5, and 1 µg kg(-1). The remifentanil doses were plotted against V(nadir)/V(preinj) in a log-probit diagram to determine what amount gave the same maximum ventilatory depression as the fentanyl dose. RESULTS: V(nadir) was [median (inter-quartile range)] 51 (38-64)% of V(preinj) after fentanyl, and 70 (61-77), 50 (46-56), and 29 (24-38)%, respectively, after remifentanil. The nadir occurred 5.0 (4.4-7.0) min after fentanyl, and 3.8 (2.7-4.6), 2.9 (2.7-3.2), and 3.0 (2.7-3.2) min after remifentanil injection. PE'CO2 at ventilation nadir was 6.26 (5.98-6.62) kPa after fentanyl, and 6.18 (6.12-6.50), 6.11 (5.91-6.45), and 6.11 (5.93-6.45) kPa after remifentanil 0.25, 0.5, and 1 µg kg(-1), respectively. A remifentanil dose of 0.47 (0.42-0.62) µg kg(-1) was equidepressant to 1 µg kg(-1) of fentanyl. Fifteen minutes after fentanyl injection, the median minute ventilation was 30-40% less than after injection of remifentanil, 0.25 and 0.5 µg kg(-1) (P<0.05). CONCLUSIONS: Fentanyl, 1 µg kg(-1), and remifentanil, 0.5 µg kg(-1), gave similar maximum ventilatory depression. The onset of and recovery from ventilatory depression were faster with remifentanil.

How common is severe pulmonary hypertension after pediatric cardiac surgery?

BACKGROUND: Pulmonary hypertension may result in significant morbidity and mortality after pediatric cardiac surgery. The objective of this study was to determine the incidence and outcome of severe pulmonary hypertension, defined as a ratio of pulmonary to systemic arterial pressure equal to or greater than 1.0, after cardiac surgery in children. METHODS: Data from all children younger than 18 years who had undergone cardiothoracic surgery from January 1, 1994, to December 31, 1998, were examined. To find children with severe pulmonary hypertension, we reviewed intensive care unit charts from patients who had been monitored with a pulmonary artery catheter after the operation (n = 151), had received mechanical ventilation for more than 4 days after the operation (n = 124), or had died in the operating room or the intensive care unit (n = 22). Intraoperative and postoperative measurements of mean pulmonary arterial pressure and postoperative echocardiographic studies during the first 3 postoperative days were used to select the children. RESULTS: During the study period, 1349 children (including 164 neonates and 511 infants, median age 12 months) underwent cardiac operations with an overall perioperative mortality of 22 patients (1.6%). Twenty-seven children (2%, median age 4.2 months) had severe pulmonary hypertension. Of these, 2 (7.4%) died within 30 days of the operation, and 3 others (11%) died within a year (median follow-up 53 months). Nitric oxide inhalation was used in 5 of the 27 cases, and it probably saved the life of 1 patient, may have helped in 1 case, and had no discernible effect in 3 cases. Severe pulmonary hypertension was most common after correction of complete atrioventricular septal defects (14%, n = 12/85). Thirteen of 131 children with Down syndrome (9.9%) had severe pulmonary hypertension. CONCLUSION: Severe postoperative pulmonary hypertension occurred after 2% of the cardiac procedures and in most cases was managed successfully with conventional treatment and had a favorable postoperative outcome. The low incidence relative to previous reports may reflect the benefits of early correction and improved intraoperative and postoperative care.

Airway closure in anesthetized infants and children: influence of inspiratory pressures and volumes.

BACKGROUND: Cyclic opening and closing of lung units during tidal breathing may be an important cause of iatrogenic lung injury. We hypothesized that airway closure is uncommon in children with healthy lungs when inspiratory pressures are kept low, but paradoxically may occur when inspiratory pressures are increased. METHODS: Elastic equilibrium volume (EEV) and closing capacity (CC) were measured with a tracer gas (SF(6)) technique in 11 anesthetized, muscle-relaxed, endotracheally intubated and artificially ventilated healthy children, aged 0.6-13 years. Airway closing was studied in a randomized order at two inflation pressures, +20 or +30 cmH(2)O, and CC and CC/EEV were calculated from the plots obtained when the lungs were exsufflated to -20 cmH(2)O. (CC/EEV >1 indicates that airway closure might occur during tidal breathing). Furthermore, a measure of uneven ventilation, multiple breath alveolar mixing efficiency (MBAME), was obtained. RESULTS: Airway closure within the tidal volume (CC/EEV >1) was observed in four and eight children (not significant, NS) after 20 and 30 cmH(2)O inflation, respectively. However, CC(30)/EEV was >CC(20)/EEV in all children (P< or = 0.001). The MBAME was 75+/-7% (normal) and did not correlate with CC/EEV. CONCLUSION: Airway closure within tidal volumes may occur in artificially ventilated healthy children during ventilation with low inspiratory pressure. However, the risk of airway closure and thus opening within the tidal volume increases when the inspiratory pressures are increased.

Patients with congenital heart malformations for noncardiac surgery.

Most patients with CHD can be safely anesthetized with regular techniques. Preoperative consultation with appropriate specialists and a well planned anesthetic management is important.

Hemodynamic effects of tracheal and intravenous adrenaline in infants with congenital heart anomalies.

BACKGROUND: If intravenous access cannot be accomplished during cardiopulmonary resuscitation in children, tracheal administration of 100 micrograms/kg of adrenaline (ten times greater than the intravenous dose) is recommended. METHODS: In a randomized crossover study we recoreded the hemodynamic effect of a low dose of intravenous adrenaline and a ten times greater tracheal dose. While anesthetized for open heart surgery, fourteen infants received one dose of adrenaline intravenously (0.3 microgram/kg) and the other tracheally (3 micrograms/kg). RESULTS: During the first 5 minutes after administration mean arterial pressure (MAP) and heart rate (HR) increased after both intravenous and tracheal administration (P < 0.001). The maximum increase in MAP was 28% (17-68%, median and range) after intravenous injection and 20% (6-69%, P < 0.05 when compared to intravenous injection) after tracheal instillation. In four infants, MAP increased less than 10% after tracheal instillation. The maximum increases in MAP and HR occurred 1 min (1-2 min) after intravenous injection and 3 min (2-4 min) after tracheal instillation (P < 0.001). CONCLUSION: Tracheal administration 3 micrograms/kg adrenaline increased mean arterial blood pressure in infants with congenital cardiac anomalies, but the increase occurred later and was less consistent than after 0.3 microgram/kg of adrenaline given intravenously.

Circulatory effects of hypoxia, acute normovolemic hemodilution, and their combination in anesthetized pigs.

BACKGROUND: Because hemodilution decreases the oxygen-carrying capacity of blood, it was hypothesized that severe hemodilution would decrease the tolerance to alveolar hypoxia. METHODS: Hemodynamics, oxygen transport, and blood lactate concentrations were compared in ten pigs with normal hematocrit (33 +/- 4%), and ten hemodiluted pigs (hematocrit 11 +/- 1%; mean +/- SD) anesthetized with ketamine-fentanyl-pancuronium during stepwise decreases in inspired oxygen fraction (FIO2; 1.0, 0.35, 0.21, 0.15, 0.10, 0.05). RESULTS: Median systemic oxygen delivery (DO2SY) became critical (the DO2SY value when arterial lactate exceeded 2.0 mmol.l-1) at 10.4 ml.kg-1.min-1 (range 6.9-16.1) in hemodiluted animals and at 11.8 ml.kg-1.min-1 (5.9-32.2) in animals with normal hematocrits (NS). The relationship between mixed venous oxygen saturation and arterial lactate values was less consistent and median critical mixed venous oxygen saturation was higher (P < 0.05) in the hemodiluted group (35%, range 21-64), than in animals with normal hematocrits (21%, 7-68%). In animals with normal hematocrit, decreasing FIO2 from 1.0 to 0.10 resulted in a decrease in DO2SY from 26.3 +/- 9.1 to 9.3 +/- 3.9 ml.kg-1.min-1 (P < 0.01). Cardiac output did not change, systemic oxygen extraction ratio increased from 0.23 +/- 0.08 to 0.68 +/- 0.13 (P < 0.01), and arterial lactate from 0.9 +/- 0.2 to 3.4 +/- 3.0 mmol.l-1 (P < 0.05). Cardiac venous blood flow, as measured by retrograde thermodilution, increased from 5.7 +/- 2.9 to 12.6 +/- 5.7 ml.kg-1.min-1 (P < 0.01). When FIO2 was reduced to 0.05, three animals became hypotensive and died. In the second group, hemodilution increased cardiac output and systemic oxygen extraction ratio (P < 0.01). Cardiac venous blood flow increased from 4.1 +/- 1.7 to 9.8 +/- 5.1 ml.kg-1.min-1 (P < 0.01), and cardiac venous oxygen saturation from 22 +/- 5 to 41 +/- 10% (P < 0.01). During the subsequent hypoxia, cardiac output and DO2SY were maintained until FIO2 = 0.15 (DO2SY = 10.1 +/- 3.3 ml.kg-1.min-1). Cardiac venous blood flow was then 18.5 +/- 10.7 ml.kg-1.min-1 (P < 0.01), but in spite of this, myocardial lactate production occurred. At FIO2 = 0.10 (DO2SY = 7.7 +/- 3.0 ml.kg-1.min-1), arterial lactate concentration increased to 8.5 +/- 2.3 mmol.l-1 (P < 0.01), and most animals became hypotensive. All hemodiluted animals died when FIO2 was decreased to 0.05 (P < 0.01 when compared to animals with normal hematocrit). CONCLUSIONS: Systemic and myocardial lactate production occurred at similar systemic oxygen delivery rates in hemodiluted and nonhemodiluted animals. Mixed venous oxygen saturation may be a less reliable indicator of inadequate oxygen delivery during hemodilution.

Pressure-volume relations of the respiratory system in healthy children.

Static pressure-volume (P-V) curves of the respiratory system were obtained in 48 healthy children (1 mo to 16 yr of age) during anesthesia and muscle paralysis. The lungs were inflated to a pressure of 25 to 40 cm H2O, and during the subsequent deflation an interrupter placed in the airway tubing opened and closed every 0.16 s. Airway flow was integrated to obtain the volume decrement between consecutive flow interruptions. Airway pressure was measured during interruptions, and a curve relating pressure to lung volume was plotted, assuming the lung volume at zero pressure to equal functional residual capacity (FRC). FRC was measured using tracer gas washout. The maximum slope of the P-V curve (maximum compliance = Crsmax, ml/cm H2O) was closely related to length (in centimeters) of the child: Crsmax = 7.7 x 10(-4) x length2.38; r = 0.97. The pressure coinciding with Crsmax was 6 +/- 1 cm H2O (mean +/- SD) in infants (1 to 6 mo of age) and 12 +/- 1 cm H2O in older children (> 1.5 yr of age). Total lung capacity (TLC) per kg body weight increased with age and was 52 +/- 13 ml/kg in infants and 87 +/- 11 mg/kg in older children. The FRC/TLC ratio was greater in infants (38 +/- 4%) than in older children (30 +/- 5%). The lung volume coinciding with Crsmax was nearly the same at all ages, when expressed as a percentage of TLC: 62 +/- 3%. Specific compliance of the respiratory system, that is, Crsmax/TLC, decreased with growth and was 0.044 +/- 0.006 cm H2O-1 in infants and 0.035 +/- 0.004 cm H2O-1 in older children. It is concluded that although the P-V relations of the respiratory system changed markedly with growth, especially during the first year of life, the lung volume (%TLC) at which maximum compliance occurred varied little.

Dissolving methohexital in a lipid emulsion reduces pain associated with intravenous injection.

Pain often accompanies intravenous injection of 1% methohexital. The aim of the present study was to test whether pain on injection could be reduced by dissolving methohexital in a lipid emulsion (study A) and whether this would affect anesthetic potency (study B). In study A, 24 healthy volunteers, 36 +/- 1 yr (mean +/- SE), were given 1 ml 1% methohexital in saline, 1 ml 1% methohexital in lipid emulsion, and 5 ml 0.1% methohexital in saline in random order. The injections were given in a small vein in the forearm at 5-min intervals. One minute after each injection, the subject was asked to assess the injection pain on a visual analog scale (0-100 mm). The pain score (median [range]) was 44.5 (0-77) after 1% methohexital in saline, 0.5 (0-26) after 1% methohexital in a lipid emulsion, and 1.0 (0-26) after 0.1% methohexital in saline. The pain score for 1% methohexital in saline was significantly greater than those for the other two solutions (P less than 0.001 for each comparison). In study B, 42 patients, 41 +/- 3 yr, were given 1% methohexital in lipid emulsion (n = 22) or 1% methohexital in saline (n = 20). A bolus of either solution was administered over 10 s, and the patient was considered asleep if there was no gross movement or response to verbal command 40-70 s after injection.(ABSTRACT TRUNCATED AT 250 WORDS)

Functional residual capacity in anesthetized children: normal values and values in children with cardiac anomalies.

To assess the increase in functional residual capacity (FRC) with growth, FRC was measured after induction of anesthesia in two groups of children. One group consisted of 74 children, 0.1-11.2 yr of age, without signs of cardiorespiratory disease (referred to here as "normal" children), and the other of 21 children, 0.2-6.9 yr of age, with cardiac malformations. Anesthesia was maintained with halothane in the normal children and with fentanyl, droperidol, and nitrous oxide in the children with cardiac anomalies. All patients were paralyzed, their tracheas intubated, and their lungs mechanically ventilated. FRC was measured with an automated tracer gas washout technique. In 70 patients the measurements were performed in duplicate with a mean coefficient of variation of 2.0%. FRC correlated significantly with height, weight, and age in both groups. Multiple regression analysis for both groups considered together indicated no significant improvement when factors for the sex of the child or for the presence of cardiac anomalies were incorporated into the model. In normal children the simple linear and nonlinear regression equations for FRC (in milliliters) versus height (in centimeters) were: FRC = -529 + 9.48 x height, r = 0.96; and FRC = 0.00175 x height2.66, r = 0.97, respectively. The corresponding equations for FRC (in milliliters) versus weight (in kilograms) were: FRC = -92 + 29.9 x weight, r = 0.93; and FRC = 9.51 x weight1.31, r = 0.95.(ABSTRACT TRUNCATED AT 250 WORDS)

Postoperative emesis after pediatric strabismus surgery: the effect of dixyrazine compared to droperidol.

Sixty-one children, ASA physical status I, aged 2-14 years, admitted for strabismus surgery were studied. All were premedicated with diazepam and atropin rectally. Anesthesia was induced with thiopental or with halothane on a facemask, and succinylcholine was given to facilitate tracheal intubation. Anesthesia was maintained with halothane and nitrous oxide. Each child was randomly assigned to receive either no antiemetic prophylaxis (control), droperidol 0.075 mg/kg, or dixyrazine 0.25 mg/kg. The drugs were injected intravenously at the end of surgery. The incidence of vomiting during the following 24 h was 65% in the control group, 48% in the droperidol group, and 25% in the dixyrazine group (P less than 0.05 as compared to the control group). Four hours after the operation, six children in the droperidol group and none in the dixyrazine group (P less than 0.05) were difficult to arouse. It is concluded that dixyrazine reduces the incidence of postoperative vomiting without causing heavy sedation.

Thiopental requirements for induction of anesthesia in neonates and in infants one to six months of age.

The authors determined the thiopental dose needed for satisfactory induction in ten neonates, 0-14 days of age, and 20 infants, 1-6 months of age. A single iv bolus of thiopental was given. Thirty seconds after injection the anesthesia mask was applied and the response was observed during the following 30 s while the patient breathed oxygen. Induction was considered satisfactory if there were no gross movements or coughing. The dose required for satisfactory induction in 50% of patients, ED50 (+/- SE), was 3.4 +/- 0.2 mg/kg in neonates and 6.3 +/- 0.7 mg/kg in infants (P less than 0.001). It is concluded that the thiopental dose needed for satisfactory induction is less in neonates than in infants.

Lung mechanics during upper abdominal surgery.

Functional residual capacity (FRC) and breath-by-breath compliance of the respiratory system (Crs) were studied after induction of anaesthesia, after insertion of retractors and after wound closure in patients undergoing upper abdominal surgery via a subcostal (n = 8) or a midline (n = 8) incision. After anaesthesia induction the mean FRC was 1.6 +/- 0.3 l. In the subcostal incision group FRC did not change between the studied stages, but Crs fell after retractor placement from 51 +/- 3 to 43 +/- 5 ml/cmH2O (p less than 0.01). In the midline incision group FRC rose by 21% (p less than 0.01) when the retractors were inserted, but regained outset level after wound closure. Crs in this group did not change significantly after retraction, but after closure of the wound it fell to 44 +/- 6 ml/cmH2O, i.e. less (p less than 0.05) than the outset value (52.6 ml/cmH2O). FRC thus did not decrease in either group, but Crs fell by about 15%. The authors conclude that the known difference in postoperative pulmonary complications between midline vs. subcostal incisions is not caused by the studied intraoperative events.

Lung function in the supine and lateral decubitus positions in anaesthetized infants and children.

We have measured dynamic lung compliance or static lung thorax compliance, functional residual capacity (FRC), and two indices of pulmonary gas mixing (pulmonary clearance delay (PCD) and single breath alveolar mixing efficiency (SBAME)) in 25 children in the supine and lateral decubitus position during nitrous oxide-halothane anaesthesia. Fifteen children (5 month-8 yr) breathed spontaneously and 10 (4 month-9 yr) underwent mechanical ventilation. Tidal volume and rate of ventilation were, respectively, 3.5-6.6 ml kg-1 and 22-46 b.p.m. in spontaneously breathing supine children, and 8.3-15 ml kg-1 and 20-30 b.p.m. in mechanically ventilated supine children, and did not differ significantly in the lateral position. There was no significant change in compliance when the child was turned to the lateral position, but FRC increased from 22 (SD7) to 25 (8) ml kg-1 (P less than 0.01) in the spontaneously breathing group and from 19 (6) to 24 (8) ml kg-1 (P less than 0.01) in the other group. In spontaneously breathing children, PCD and SBAME indicated a somewhat impaired pulmonary gas mixing (P less than 0.05) after the child had been turned to the lateral position, but no change occurred in the other group. These findings suggest that the distribution of ventilation in anaesthetized children in the lateral position is similar to that reported previously in anaesthetized adults.

Ventilation inhomogeneity during controlled ventilation. Which index should be used?

Six indexes for diagnosing uneven ventilation by tracer gas washout were studied. The indexes were lung clearance index, mixing ratio, Becklake index, multiple-breath alveolar mixing inefficiency, moment ratio, and pulmonary clearance delay, all of which increase with impaired pulmonary gas mixing. In model lung tests, indexes that compared the actual washout curve with a calculated ideal curve (mixing ratio, multiple-breath alveolar mixing inefficiency, and pulmonary clearance delay) were unaffected by changes in tidal volume and series dead space, whereas the others varied markedly. In both spontaneously breathing and mechanically ventilated patients all indexes showed a significant difference between smokers and nonsmokers (P less than 0.002), but the indexes were somewhat different in their assessment of different ventilatory patterns. However, the mean value for all indexes, with the exception of mixing ratio, was smallest with a fast insufflation followed by an end-inspiratory pause. Any of the indexes may be useful if its limitations are recognized, but mixing ratio, multiple-breath alveolar mixing inefficiency, and pulmonary clearance delay seem preferable, because they are not affected by changes in tidal volume and dead space fraction.

Measurement of lung volume by sulfur hexafluoride washout during spontaneous and controlled ventilation: further development of a method.

An open circuit tracer gas washout method for measurement of lung volume in patients during anesthesia and intensive care is described and tested. The method employs a device for dispensing the tracer gas, sulfur hexafluoride (SF6), a fast SF6 analyzer, a pneumotachograph, and a computer. The dispensing device delivers SF6 into the airway in proportion to instantaneous inspiratory flow so that inspiratory SF6 concentration is held constant, usually at about 0.5%, regardless of the inspiratory flow pattern. The amount of SF6 present in the lungs at the end of a washin is calculated during washout from signals representing expired SF6 concentration and expired flow. From this, lung volume is derived. Accurate and reproducible results were obtained in lung model tests during ventilation with air, N2O in O2, and halothane in O2. Functional residual capacity (FRC) was measured both with SF6 washout and nitrogen washout in five mechanically ventilated patients. This gave the regression equation: FRCSF6 = 10 ml + 1.04 x FRCN2, r = 0.99. A similar close agreement was observed for total lung capacity (TLC) and residual volume (RV) measurements in eight healthy, spontaneously breathing subjects: TLCSF6 = 91 ml + 1.01 x TLCN2, r = 0.99; RVSF6 = -32 ml + 0.97 x RVN2, r = 0.95. Comparison with body plethysmography in eight healthy, sitting subjects gave the regression equation: FRCSF6 = 180 ml + 0.96 x FRCbox, r = 0.99. The median (range) for the coefficient of variation at duplicate determinations in 10 anesthetized, paralyzed, and mechanically ventilated adults was 3.0% (0.2-6.6%).(ABSTRACT TRUNCATED AT 250 WORDS)

Ventilatory consequences of the lateral position and thoracotomy in children.

Functional residual capacity (FRC), breath-by-breath compliance of the respiratory system (Crs) and arterial oxygen tension (PaO2) were measured in ten children, two months to nine years of age, during anaesthesia for surgical correction of patent ductus arteriosus or coarctation of the aorta. The children were mechanically ventilated with halothane, nitrous oxide and oxygen. FIO2 was kept constant in each child. After induction of anaesthesia, FRC was 17 +/- 7 ml X kg-1 (mean +/- 1 SD), corresponding to 60 +/- 22 per cent of a predicted awake value. FRC increased to 21 +/- 8 ml X kg-1 (p = 0.0005) when the child was turned to its right side and decreased to 13 +/- 5 ml X kg-1 (p = 0.0003) when the pleura was opened. No significant change in Crs or PaO2 occurred during these manoeuvres. Retraction of the upper lung to visualize the great vessels caused a significant decrease in FRC, Crs, and PaO2. The lowest PaO2 observed during this stage was 70.0 mmHg. After surgery FRC and PaO2 were about the same as before surgery while Crs had decreased from 0.87 +/- 0.18 preoperatively to 0.64 +/- 0.15 ml X cmH2O-1 X kg-1 (p = 0.0069). This study shows that FRC increases when mechanically ventilated children are placed in the lateral position, and that thoracotomy is associated with marked changes in FRC, Crs and PaO2.

Changes in lung volume and lung-thorax compliance during cardiac surgery in children 11 days to 4 years of age.

To examine the effects of cardiac surgery and cardiopulmonary bypass (CPB) on the lung, functional residual capacity (FRC) and lung-thorax compliance were measured at four stages during open heart surgery in 15 children. The patients were anesthetized with fentanyl/droperidol and N2O/O2, paralyzed, and ventilated with volume-controlled mechanical ventilation at 20-30 breaths/min. FRC was measured by tracer gas washout. Static lung-thorax compliance (CLT) was calculated as tidal volume divided by the airway pressure difference between the end of the postinspiratory pause and the end of the expiration, and also from the increase in FRC caused by adding 5 cmH2O of PEEP (CLT[FRC]). Before skin incision, both FRC and compliance were closely correlated with weight and length. During this stage, FRC was 21 +/- 5 ml/kg, CLT 0.90 +/- 0.21, and CLT(FRC) 1.28 +/- 0.35 ml X cmH2O-1 X kg-1 X PEEP 5 increased FRC by 34 +/- 9%. In patients with intact pleural cavities throughout the operation (n = 10), FRC increased by 4 +/- 2 ml/kg when the sternum was retracted (P less than 0.01). During CPB, FRC decreased by 4 +/- 3 ml/kg (P less than 0.01), and FRC at the end of surgery was 5 +/- 4 ml/kg less than before skin incision (P less than 0.01). In these ten children, there was a 13% and 6% decrease in mean CLT and CLT(FRC), respectively, during the operation (P less than 0.05) and mean CLT(FRC) was at least 40% greater than CLT during all four stages (P less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)

Changes in functional residual capacity during cardiac surgery.

A gas washout technique was used to measure the functional residual capacity (FRC) in eight patients during anaesthesia for cardiac surgery. The patients were anaesthetized with droperidol, fentanyl and nitrous oxide, alcuronium was given and the lungs were ventilated with a volume controlled ventilator. FRC was measured at three stages before skin incision, after sternotomy but before cardiopulmonary bypass, and after closure of the sternum. The pleural cavities were intact in all patients during the operation. FRC before skin incision was 1.7 +/- 0.5 litre (mean +/- 1 SD). A 55% mean increase in volume was noted after sternotomy and placement of the sternal retractor (P less than 0.001). Mean FRC after sternal closure was 16% lower than the preincision value (P less than 0.05). Arterial Po2 was measured in 22 other patients who underwent coronary artery bypass surgery and in whom F/o2 was 0.5. Pao2 increased significantly when the sternum was opened, but decreased after cardiopulmonary bypass. There was a further significant decrease on closure of the sternum.