Effects of inspired gas composition during anaesthesia for abdominal hysterectomy on postoperative lung volumes

Transcription

1 British Journal of Anaesthesia 1995; 75: Effects of inspired gas composition during anaesthesia for abdominal hysterectomy on postoperative lung volumes C. J. JOYCE AND A. B. BAKER Summary We have studied 51 patients who were allocated randomly and prospectively to receive either 100 % oxygen (n 16), 70 % nitrous oxide in oxygen (n 18) or 30 % oxygen in nitrogen (n 17) as the inspired gas during anaesthesia for abdominal hysterectomy. Lung volumes were measured before and after surgery. TLC, VC, FVC and FEV 1 but not RV or FRC were reduced after surgery. There were no significant differences between the three treatment groups in any of the lung volumes measured. We conclude that absorption atelectasis during anaesthesia is not the main cause of perioperative changes in lung volume after abdominal hysterectomy. Any effect of the inspired gas is likely to be of limited clinical significance. (Br. J. Anaesth. 1995; 75: ) Key words Surgery, gynaecological. Anaesthetics gases, nitrous oxide. Lung, volume. Lung, atelectasis. Atelectasis and impairment of respiratory function are common problems after surgery. While the effects of the inspired gas composition during anaesthesia on postoperative respiratory function have been examined in patients undergoing cardiac [1] and upper abdominal [2] surgery, these effects have not been examined in patients undergoing lower abdominal surgery. The size and site of incision have a major effect on postoperative respiratory function [3, 4]. The changes in respiratory function in patients undergoing lower abdominal cavity surgery are relatively small [5, 6]. There are no studies documenting the changes in respiratory function after abdominal hysterectomy via a lower abdominal transverse incision, but gynaecological procedures via a vertical subumbilical midline incision decreased functional residual capacity (FRC) to 88 % of preoperative values [6]. Alveolar atelectasis may be caused by a variety of factors including compression, loss of surfactant, and absorption of gas [7], of which the latter is most likely. During anaesthesia, high inspired oxygen concentrations are administered routinely to avoid hypoxaemia. Lung volume is reduced by anaesthesia, and evidence of atelectasis after short periods of oxygen breathing has been demonstrated in normal conscious subjects under a variety of conditions that force them to breathe at low lung volume [8 12]. Airways closure may occur during anaesthesia [13 15], particularly in middle aged or elderly patients [13]. The rate of absorption of gas from unventilated areas of lung depends on the composition of the inspired gas [16 19]. When the inspired gas is 100 % oxygen, the rate of collapse is faster than when air oxygen mixtures are breathed [17, 18]. When the inspired gas is a mixture of nitrous oxide and oxygen and FI O 2 0.3, collapse is more rapid than when 100 % oxygen is breathed [18]. If absorption atelectasis during anaesthesia is an important mechanism in the genesis of pulmonary collapse after surgery, then using air oxygen mixtures as the inspired gas should reduce the amount of postoperative pulmonary collapse. We have examined the changes in lung volumes that occur after abdominal hysterectomy with a lower abdominal transverse incision, and tested the null hypothesis that the changes in lung volume after abdominal hysterectomy do not depend on whether ventilation during operation is with 100 % oxygen, a nitrous oxide oxygen mixture or an air oxygen mixture. Patients and methods To estimate the number of subjects required for this study, we examined the literature on the changes in lung volume after lower abdominal surgery [6]. Power analysis [20] performed on these data indicated that 20 patients would be required in each group. The study was approved by the Ethics Committee of the Otago Area Health Board. We studied patients undergoing elective abdominal hysterectomy via a transverse lower abdominal incision. Exclusion criteria included a large intra-abdominal mass, the presence of obstructive lung disease (FEV 1/FVC ration 70 % in the sitting position), a history of respiratory disease or weight 110 kg. Patients were assessed the day before surgery. Forced expiratory volume in 1 s (FEV 1) and forced C. J. JOYCE*, MB, CHB, PHD, FANZCA, A. B. BAKER, MB, BS, DPHIL, FANZCA, FRCA, Department of Anaesthesia and Intensive Care, University of Otago, Dunedin, New Zealand and Dunedin Public Hospital, Dunedin, New Zealand. Accepted for publication: February 27, Present addresses: *Department of Intensive Care, Royal Melbourne Hospital, Victoria 3050, Melbourne, Australia. Department of Anaesthetics, University of Sydney, Royal Prince Alfred Hospital, Sydney, Australia.

2 418 British Journal of Anaesthesia vital capacity (FVC) were measured in the seated position to assess suitability for inclusion. Subjects included in the trial were allocated randomly and prospectively to either the oxygen group, the nitrous oxide group or the air group. Control measurements of FRC, residual volume (RV), vital capacity (VC), total lung capacity (TLC), FEV 1, FVC and visual analogue pain score (VAS) were made in a seated position (see [1] for methods of measurement). Patients were premedicated with temazepam 20 mg approximately 1 h before surgery. Anaesthesia was induced with thiopentone 4 mg kg 1 and morphine 0.15 mg kg 1. Vecuronium 0.1 mg kg 1 with incremental doses as required was given for neuromuscular block. After tracheal intubation, the lungs were ventilated with a tidal volume of 10 ml kg 1, with the rate adjusted to maintain end-tidal Pco 2 at kpa via an HME humidifier and circle circuit with a fresh gas flow of 6 litre min 1. I O 2 F in the oxygen group was maintained at 1.0, in the nitrous oxide group at 0.3 with the balance nitrous oxide and in the air group at 0.3 with air and oxygen. In the oxygen and air groups, isoflurane was administered with the vaporizer set initially at 2 %, while in the nitrous oxide group the vaporizer was set at 1 %. Isoflurane doses were varied as required according to clinical signs. At the end of surgery residual neuromuscular block was antagonized with atropine 1.2 mg and neostigmine 2.5 mg, the trachea extubated and the patient taken to the recovery ward. While in recovery patients received 26 % oxygen, until S p O on air was 2 95 %. Postoperative analgesia comprised morphine 1 4 mg h 1 i.v. with increments as required, and diclofenac 50 mg three times a day orally or rectally. The morphine infusion was discontinued when the nursing staff and patient felt that opioids were no longer necessary. Respiratory function was measured on the morning of postoperative days 1, 2 and 3 in the seated position using the same technique as before. All tests were performed by the same investigator who was unaware of the treatment group, as were the patient, and the nursing and medical staff. When a subject was unable or unwilling to complete any set of respiratory function tests, no additional measurements of respiratory function were performed for the purposes of the study. Statview for the MacIntosh version 4.0 was used for statistical analysis. The first step was to assess which lung volumes changed significantly in the perioperative period. For each lung volume a comparison of the measurement on the 4 study days was made with one-way repeated measures analysis of variance, using all 51 subjects who completed the study. The within-subjects factor was the day on which the measurement was made (preoperative, postoperative day 1, postoperative day 2, postoperative day 3). Lung volumes that did not change significantly with operation (RV and FRC) were not examined further. Summary measures were used to analyse the serial measurements of lung volume in the three treatment groups, as recommended by Matthews and col- leagues [21]. For a given summary measure there was a single value for each subject, allowing comparison between the three treatment groups with ANOVA. The summary measure analysed for each lung volume was the maximum volume recorded for each subject on any of the three postoperative measurements, where volume (preoperative volume current volume)/preoperative volume. For each summary measure of a lung volume there was a single value for each subject, allowing comparison between the results for the three treatment groups with one-way factorial analysis of variance. Analyses were carried out on 51 subjects who completed the full measurement study. Three treatment groups were used (oxygen, nitrous oxide and air). It was decided initially to have 20 subjects in each treatment group. A difficulty in this type of clinical trial is that not all subjects entering the trial complete the full study. Because of this a pool of remaining treatments method of randomization was used [1]. During most of the study, patients undergoing abdominal hysterectomy were admitted on the day before surgery. Hospital administration policy changed during the latter part of the study, so that patients were admitted on the day of surgery. This caused major problems in recruitment and the logistics of making the preoperative measurements. Because of this the study was terminated after the full set of measurements had been completed in 16 subjects in the oxygen group, 18 in the nitrous oxide group and 17 in the air group. No significant difference was demonstrated between the treatment groups for any of the lung volumes measured. The next question asked was what difference between the mean values for the treatment groups would be detected with an acceptable type 2 error level?. Sokal and Rohlf [22] described a method for calculating the number of replications needed to detect a given true difference between means using ANOVA. This method is readily adapted to find the true difference between the means that would be detected with a given sample size, for predetermined and. Results Of the 69 patients who entered the study, we obtained complete measurements in 51. In the 69 patients who entered the trial there were no statistically significant differences (P 0.05) between the oxygen (n 20), nitrous oxide (n 24) and air Table 1 Patient data (mean (SEM) [range]) for all those who entered the study (n 69). Twenty of 69 patients had smoked in the previous 12 months Age (yr) 43 [28 48] Height (cm) (0.686) [ ] Weight (kg) 70.8 (1.7) [43 107] Duration of operation (min) 103 (3) [55 175] FRC (litre) (0.081) [ ] VC (litre) (0.071) [ ] RV (litre) (0.065) [ ] TLC (litre) (0.090) [ ] FEV1 (litre) (0.056) [ ] FVC (litre) (0.072) [ ]

3 Postoperative lung volumes and inspired gas 419 Table 2 Comparison between the treatment groups air (n 17), nitrous oxide (N 2 O) (n 18) and oxygen (O 2 ) (n 16) for maximum volume for VC, TLC, FEV 1 and FVC in patients who completed the full study. P values from factorial one-way analysis of variance. o Difference between two of the three means detected with the specified power and tot maximum volume for the three groups combined. Thus o/ tot fraction of the maximum perioperative change in that lung volume that would be detected with the specified power when it is caused by a difference between the treatment groups Air Air N 2 O N 2 O O 2 O 2 P o/ tot 80 % power o/ tot 90 % power Max VC % 85 % Max TLC % 82 % Max FEV % 70 % Max FVC % 70 % Table 3 Perioperative changes in lung volumes for subjects that completed the full study in the air, nitrous oxide (N2O) and oxygen (O2) groups. Lung volumes are expressed as a fraction of the preoperative lung volume in that patient (mean (SEM)) Group Postop. day FRC VC RV TLC FEV1 FVC Air (0.036) (0.026) (0.048) (0.022) (0.019) (0.019) Air (0.031) (0.028) (0.043) (0.025) (0.015) (0.021) Air (0.023) (0.021) (0.034) (0.016) (0.009) (0.012) O (0.042) (0.034) (0.070) (0.022) (0.017) (0.015) O (0.045) (0.022) (0.069) (0.029) (0.019) (0.017) O (0.031) (0.020) (0.046) (0.019) (0.011) (0.010) N2O (0.031) (0.024) (0.042) (0.018) (0.015) (0.015) N2O (0.027) (0.027) (0.036) (0.018) (0.018) (0.018) N2O (0.031) (0.024) (0.044) (0.018) (0.017) (0.015) groups (n 25) in variables assessed before entry into the trial (age, height, weight, smoking history), in preoperative measurements of lung volume (FRC, VC, RV, TLC, FEV 1, FVC) or in duration of operation (table 1). In the 51 patients who completed the full study, there were also no statistically significant differences between the oxygen (n 16), nitrous oxide (n 18) and air groups (n 17) in these same variables. Thus there is no suggestion that the treatment groups were obtained from different patient populations, or that bias caused by poor matching of the subjects in the three groups occurred. Comparison of these same variables between subjects that completed the trial and subjects that withdrew from the trial showed no statistically significant differences. Analysis of VAS pain scores showed no significant difference (P 0.05) between the oxygen, nitrous oxide and air groups in the before scores, after scores, or difference scores on any of the perioperative days that lung volumes were measured. This makes it unlikely that differences in pain between the groups have biased these results. While VC, FVC, TLC and FEV 1 decreased after surgery (all P 0.001), there were no significant changes (P 0.05) in RV or FRC (one-way repeated measures ANOVA). FRC and RV were therefore not examined further. There were no significant differences between the air, nitrous oxide and oxygen groups in any of the summary measures of lung volumes examined (maximum volume for VC, TLC, FEV 1 and FVC). More detail of these analyses is given in table 2. Results of lung volumes measurements are given in table 3. Power analysis indicated that the study had an 80 % power to detect an effect caused by the treatment group (air, nitrous oxide or oxygen), that is as large or larger than 75 % of the maximum perioperative change in lung volume (60 73 % depending on which lung volume is considered (see table 2), which corresponds to a difference as great or greater than 12 % of the preoperative lung volume ( % depending on which lung volume is considered)). Discussion In this study FVC, FEV 1, VC and TLC were reduced significantly after operation. There were no significant differences between the changes in FVC, FEV 1, TLC and VC measured in the three treatment groups. Power analysis presented in table 2 shows that if the changes seen in these lung volumes were caused entirely by an effect of the inspired gas mixture (i.e. absorption atelectasis), the trial would have in excess of 90 % power to detect this effect. Indeed, if only 60 % of the change in FEV 1 or FVC after operation was caused by an effect of the inspired gas, then the trial would have 80 % power to detect

4 420 British Journal of Anaesthesia this. The addition of nitrogen to the inspired gas mixture during anaesthesia did not reduce the perioperative changes in lung volume. In fact there was a trend towards a greater reduction in lung volume in the air group than in the other groups. Therefore, the changes in TLC, VC, FEV 1 and FVC cannot be accounted for by intraoperative absorption atelectasis alone. Some contribution of absorption atelectasis cannot be excluded, but it is unlikely to be clinically important. One criticism of this study is that 26 % of subjects did not complete the full study, and their withdrawal occurred after randomization. Subjects with more atelectasis could be more prone to withdraw and this could mask an effect of the inspired gas on atelectasis. This is unlikely because of the 18 patients who withdrew, four were in the oxygen group, six were in the nitrous oxide group and eight were in the air group (chi-square, P ). We have studied patients undergoing abdominal hysterectomy by a transverse lower abdominal incision. There were statistically significant decreases in VC (16 %), TLC (12 %), FVC (12 %) and FEV 1 (12 %). FRC and RV did not change. These changes are relatively small, but this is not surprising as postoperative pulmonary complications are relatively uncommon after lower abdominal operations [22, 23]. Drummond and Littlewood [6] examined changes in FRC in patients undergoing gynaecological procedures with a subumbilical midline incision. They found that FRC on the day after surgery had decreased to an average of 88 % of the value before operation. We found no significant change in FRC with operation, but a lower abdominal transverse muscle splitting incision was used, which may explain the difference. The size and site of incision have a major effect on postoperative respiratory function after abdominal surgery [3, 4]. Postoperative pulmonary collapse is a common problem after upper abdominal [24] and thoracic operations [25]. The incidence of postoperative pulmonary complications is greatest in patients undergoing upper abdominal or thoracic surgery, and becomes progressively less with lower abdominal and peripheral surgery [22, 23]. The decrease in FVC and P a O after abdominal operations is greater 2 than after peripheral operations, and is greater with upper abdominal incisions than with lower abdominal incisions [5]. The finding of relatively small changes in respiratory function after hysterectomy via a lower abdominal transverse incision is consistent with these published results. An underlying assumption of this study is that the development of atelectasis is reflected by changes in lung volume. There is good evidence that perioperative atelectasis (on CT) is related closely to lung volume, particularly FRC (see [1]). In the perioperative period lung volumes are better indices of atelectasis than chest x-rays. Under some circumstances it is possible that when some lung units collapse, others may distend, and there may be little change in FRC. Green s findings [26] suggest that this may occur at least in some patients with acceleration atelectasis. The most likely mechanism for this would be by expansion of still opened and ventilated alveoli as a result of a decreased pleural pressure created by the collapsed alveoli. Under these circumstances, however, VC and TLC are reduced as the collapsed alveoli are not available to contribute to lung volume during a full expansion. Another explanation might be that the variability of VC measurement is less than that of FRC measurement [27], and thus a study with the same numbers using FRC as an index of atelectasis will be less powerful than one using VC. Under these circumstances, however, VC and TLC are reduced as the collapsed alveoli are not available to contribute to lung volume during a full expansion. The lack of change in FRC and RV with operation in this study does not therefore exclude atelectasis as a possible cause for the alterations in VC, FEV 1, FEV and TLC. Nevertheless, in the postoperative period other factors such as pain, and muscle weakness or spasm are likely contributing factors to the alterations in VC. The lung volumes which showed significant changes after operation are those which are most affected by pain. Given the relatively minor changes in lung volumes it may be that they are caused entirely by pain, and regional analgesia might entirely abolish them. This study shows that the changes are not caused entirely by intraoperative absorption atelectasis. Acknowledgements This work was funded in part by the Australia and New Zealand College of Anaesthetists. We acknowledge the assistance of P. Herbison (biostatistician, University of Otago Medical School) in the design and statistical analysis of this trial. References 1. Joyce CJ, Baker AB, Chartres S. Influence of inspired nitrogen concentration during anaesthesia for coronary artery bypass grafting on postoperative atelectasis. British Journal of Anaesthesia 1995; 75: Logan DA, Spence AA, Smith G. Postoperative pulmonary function. A comparison of ventilation with nitrogen or nitrous oxide during anaesthesia. Anaesthesia 1977; 32: Hedenstierna G. Mechanisms of postoperative pulmonary dysfunction. Acta Chirugica Scandinavica 1988; (Suppl. 550): Sydow FW. The influence of anaesthesia and postoperative analgesic management on lung function. Acta Chirurgica Scandinavica 1988; (Suppl. 550): Diament ML, Palmer KNV. Postoperative changes in gas tensions of arterial blood and in ventilatory function. Lancet 1966; 2: Drummond GB, Littlewood DG. Respiratory effects of extradural analgesia after lower abdominal surgery. British Journal of Anaesthesia 1977; 49: Rahn H, Farhi LE. Gaseous environment and atelectasis. Federation Proceedings 1963; 22: Baker AB, Restall R. Changes in residual volume following oxygen breathing. British Journal of Anaesthesia 1983; 55: Baker AB, McGinn A, Joyce CJ. Effect on lung volumes of oxygen concentration when breathing is restricted. British Journal of Anaesthesia 1993; 70: Dahlbäck GO, Balldin UI. Positive pressure oxygen breathing and pulmonary atelectasis during immersion. Undersea Biomedical Research 1983; 10: Nunn JF, Williams IP, Jones JG, Hewlett AM, Hulands GH, Minty BD. Detection and reversal of pulmonary absorption collapse. British Journal of Anaesthesia 1978; 50: Tacker WA, Balldin UI, Burton RR, Glaister DH, Gillingham KK, Mercer JR. Induction and prevention of ac-

5 Postoperative lung volumes and inspired gas 421 celeration atelectasis. Aviation, Space, and Environmental Medicine 1987; 58: Bergman NA, Tien YK. Contribution of the closure of pulmonary units to impaired oxygenation during anesthesia. Anesthesiology 1983; 59: Don HF, Wahba WM, Craig DB. Airway closure, gas trapping, and the functional residual capacity during anesthesia. Anesthesiology 1972; 36: Weenig CS, Pietak S, Hickey RF, Fairley HB. Relationship of preoperative closing volume to function residual capacity and alveolar arterial oxygen difference during anesthesia with controlled ventilation. Anesthesiology 1974; 41: Coryllos PN, Birnbaum GL. Studies in pulmonary gas absorption in bronchial obstruction. American Journal of the Medical Sciences 1932; 183: Dale WA, Rahn H. Rate of gas absorption during atelectasis. American Journal of Physiology 1952; 170: Joyce CJ, Baker AB, Kennedy RR. Gas uptake from an unventilated area of lung: computer model of absorption atelectasis. Journal of Applied Physiology 1993; 74: Lichtheim L. Versuche über Lungenatelektase. Archiv für Experimentellen Pathologie und Pharmakologie 1879; 10: Sokal RR, Rohlf FJ. Biometry. The Principles and Practise of Statistics in Biological Research, 1st Edn. San Francisco: WH Freeman, Matthews JNS, Airman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. British Medical Journal 1990; 300: Pooler HE. Relief of post-operative pain and its influence on vital capacity. British Medical Journal 1949; Wightman JAK. A prospective survey of the incidence of postoperative pulmonary complications. British Journal of Surgery 1968; 55: Goodman LR. Postoperative chest radiograph: 1. Alterations after abdominal surgery. American Journal of Roentgenology 1980; 134: Benjamin JJ, Cascade PN, Rubenfire M, Wajszczuk W, Kerin NZ. Left lower lobe atelectasis and consolidation following cardiac surgery: The effect of topical cooling on the phrenic nerve. Radiology 1982; 142: Green ID. The Pathogenesis and Physiological Effect of Pulmonary Collapse Induced by Breathing Oxygen and Increased Gravitational Force. London: London University, Hruby J, Butler J. Variability of routine pulmonary function tests. Thorax 1975; 30: