Effect of epidural bupivacaine vs combined epidural bupivacaine and morphine on gastrointestinal function and pain after major gynaecological surgery

British Journal of Anaesthesia 87 (5): 727±32 (2001)

Effect of epidural bupivacaine vs combined epidural bupivacaine and morphine on gastrointestinal function and pain after major gynaecological surgery H. Jùrgensen1*, J. S. Fomsgaard1, J. Dirks1, J. Wetterslev1, B. Andreasson2 and J. B. Dahl1 1

Department of Anaesthesiology and Intensive Care and 2Department of Obstetrics and Gynaecology, Herlev University Hospital, Copenhagen County, Denmark *Corresponding author In a double-blind study, we investigated the effects of postoperative epidural local anaesthetic, with or without addition of epidural morphine, on postoperative pain and gastrointestinal function in patients scheduled for radical hysterectomy and pelvic lymphadenectomy. Forty patients were randomized into two study groups: 48-h postoperative epidural 0.2% bupivacaine 8 ml h±1 (bupi group) or 48-h postoperative epidural 0.2% bupivacaine/morphine 50 mg at 4 ml h±1 (bupi/morph group). Patients were observed for at least 96 h after surgery. No differences in pain at rest, during cough or mobilization were observed. Patients in the bupi group requested a signi®cant greater amount of supplementary analgesics, but times to ®rst ¯atus and defaecation were reduced compared with patients in the bupi/morph group. Itching was a signi®cant problem in patients in the bupi/morph group. No differences in postoperative nausea and vomiting, mobilization or time to discharge from hospital were observed between groups. The addition of morphine to postoperative epidural bupivacaine has only limited effect on pain relief and increases time to normalization of gastrointestinal function. Br J Anaesth 2001; 87: 727±32 Keywords: anaesthetic techniques, epidural; anaesthetics local, bupivacaine; analgesics local, bupivacaine; pain, postoperative; surgery, gynaecological Accepted for publication: June 12, 2001

Major gynaecological surgery is associated with severe postoperative pain, prolonged gastrointestinal paralysis and a high incidence of postoperative nausea and vomiting (PONV).1±6 Postoperative gastrointestinal paralysis is well known after surgery and severe trauma. Although worsened if the peritoneum is entered, the duration of surgery does not in¯uence the severity of postoperative gastrointestinal ileus.7 The pathophysiology of postoperative alterations in gastrointestinal motor function is complex and has not been clari®ed. Experimental studies show activity in spinal efferents, inhibitory non-adrenergic non-cholinergic vagal efferent nerves and in sympathetic efferent nerves, but the peptidergic neuronal systems of the enteric nervous system have still not been elucidated.6±8 Administration of epidural local anaesthetics results in effective pain relief and improved gastrointestinal motility compared with opioid-based analgesia after hysterectomy.5 6 In healthy volunteers, epidural bupivacaine alone does not affect gastric emptying or bowel function,9 10 whereas

addition of epidural opioid delays gastric emptying and prolongs orocaecal transit time.11 Only two randomized studies comparing postoperative epidural local anaesthetic with combined local anaesthetic and opioid have evaluated both gastrointestinal recovery and postoperative pain after laparotomy.12 13 In the study by Asantila and colleagues,12 analgesia with the combination of bupivacaine and morphine was superior to that with bupivacaine alone. However, gastrointestinal recovery was delayed in patients receiving the combination. In the study by Liu and colleagues,13 no signi®cant differences in either analgesia or time to ®rst postoperative ¯atus was observed between patients receiving bupivacaine and those receiving bupivacaine combined with morphine. The aim of this study was to investigate the effect of continuous thoracic epidural bupivacaine alone vs combined epidural bupivacaine and morphine on postoperative gastrointestinal function and pain after radical hysterectomy and pelvic lymphadenectomy.

Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2001

Jùrgensen et al.

Patients and methods We studied 40 patients aged 18±75 yr undergoing radical hysterectomy and pelvic lymphadenectomy through an inferior median incision. Informed written consent was obtained from all patients and approval was obtained from the local ethics committee and the Danish National Board of Health. Patients were recruited from the Department of Gynaecology, Herlev University Hospital between February 1999 and October 2000. Patients were not included if they were unable to cooperate, body weight was >[1.33body height (cm)±100], had a history of drug or alcohol abuse, chronic pain or daily intake of analgesics or had contraindications to non-steroidal anti-in¯ammatory drugs (NSAIDs) or to insertion of an epidural catheter. The study was double-blind and randomized with a computer program. The study drugs (bupivacaine 2 mg ml±1 and bupivacaine 2 mg ml±1/morphine 50 mg ml±1) were prepared by the hospital pharmacy in identical containers marked with the name of the project, the investigator's name and consecutive numbers. One hour before surgery, patients received sublingual triazolam 0.125 mg. Before induction of general anaesthesia, an 18-gauge epidural catheter was inserted through a Touhy needle and advanced 4±5 cm into the epidural space at the T10±11 level and a 4-ml test dose of lidocaine 20 mg ml±1 with epinephrine was administered. A bolus dose of 16 ml of lidocaine 20 mg ml±1 with epinephrine was administered, followed by continuous infusion of lidocaine 20 mg ml±1 at 8 ml h±1 throughout the operation. General anaesthesia was induced with propofol 1.5± 2.5 mg kg±1, alfentanil 1.0 mg and rocuronium 0.6 mg kg±1, and an endotracheal tube was passed. Propofol without nitrous oxide was given for maintenance of anaesthesia. Further alfentanil 1.0 mg was administered at the discretion of the anaesthetist. Hypotension was treated by i.v. infusion of isotonic sodium chloride, HAES 6% or ephedrine 5 mg i.v. in incremental doses when systolic blood pressure was below 90 mm Hg. Fluid and blood losses were replaced according to the prescriptions of the department. At skin closure, patients were randomized to receive an epidural bolus dose of 8 ml bupivacaine 2 mg ml±1 followed by a continuous epidural infusion of bupivacaine 2 mg ml±1 at 8 ml h±1 for 48 h (bupi group) or 4 ml bupivacaine 2 mg ml±1/morphine 50 mg ml±1 followed by a continuous epidural infusion of bupivacaine 2 mg ml±1/morphine 50 mg ml±1 at 4 ml h±1 for 48 h (bupi/morph group). Administration of the epidural bolus dose and initiation of the Deltec pump (Pharmacia) for the postoperative epidural infusion were performed by a person not involved in postoperative data collection or pain assessment. During the 48 h of infusion, the Deltec pump was covered by a box, so blinding the patient and assessors. For the ®rst 48 h after surgery, all patients received paracetamol 2 g rectally every 8 h. For the ®rst 72 h after surgery, ketorolac 15 mg was administered i.m. with a minimum interval of 8 h, on patient

demand. If additional analgesic was required before another dose of ketorolac could be administered, morphine 0.125 mg kg±1 was given i.m. In the recovery ward, morphine 0.0625 mg kg±1 i.v. was administered on patient demand. Patients were given ondansetron 4 mg i.v. with a lockout period of 8 h if they required an antiemetic. If additional antiemetic was required before another dose of ondansetron could be administered, metoclopramide 10 mg i.v. was given. Before surgery, all patients were instructed in the use of the visual analogue scale (VAS) pain score, and to request supplementary analgesics and antiemetics if needed. VAS pain scores (0 mm = no pain, 100 = worst pain imaginable) at rest, on coughing and at mobilization from the supine to the sitting position were assessed by the patients 4, 6, 24, 48, 50, 54, 72 and 96 h after surgery. The numbers of ketorolac and morphine doses were recorded at the same intervals. Levels of sensory block to pinprick bilaterally were assessed 6, 24, 48, 50 and 54 h after surgery. In the event of asymmetrical blockade, the greater spread was recorded. Motor block was assessed using a four-point modi®ed Bromage scale (0 = no motor block, 1 = inability to raise extended legs, 2 = inability to ¯ex knees, 3 = inability to ¯ex ankle joints) 6, 24, 48, 50 and 54 h after operation. Ability to walk (0 = no dif®culty, 1 = little dif®culty, 2 = very dif®cult, 3 = impossible, 4 = dif®cult or impossible for other reasons than motor block) was assessed 6, 24, 48, 50 and 54 h after surgery. The actual duration of mobilization (<1, 1±4 or >4 h) during the 24±48 and 48±72 h intervals was assessed 48 and 72 h after surgery. Patients who were not able to mobilize after >4 h were asked the primary reason for this. At all visits, gastrointestinal function was assessed by asking the patients if and when they noted ®rst ¯atus and stool, and if they had experienced nausea (0 = no, 1 = light, 2 = moderate, 3 = severe) and/or vomiting since the last assessment. Patients were encouraged to eat and drink whenever possible after the operation. Ingestion of prede®ned quantities of food and beverages was assessed at 24, 48 and 72 h. Every 24 h after surgery, patients were assessed as ready or not ready for discharge from hospital by four discharge criteria: (1) normal defaecation and no urinary retention; (2) able to mobilize and dress; (3) need for opioid; (4) surgical complication requiring hospitalization. When the patient replied `yes' to the ®rst two and `no' to the last two questions, he or she was declared ready for discharge from hospital. The primary end-point of this study was time to ®rst postoperative defaecation. Calculation of sample size was based on an unpublished pilot study performed at our institution and a previous study of postoperative gastrointestinal function after hysterectomy.12 For calculation, the following variables were used: type 1 error 5%; type 2 error 20%; minimal difference not to be overlooked, 24 h reduction in time to ®rst postoperative stool in the bupi

728

Epidural bupivacaine vs bupivacaine/morphine Table 1 Patient and perioperative characteristics [median (range) or number]. No signi®cant differences Group Bupi

Table 2 VAS pain scores (0±100 mm) at rest [median (interquartiles)] Postoperative time

Group

Bupi/morph 4 6 24 48 50 54 72 96

No. of patients at t=0 19 20 No. of patients at t=24 17 19 No. of patients at t=48 17 19 Age (yr) 41 (26±69) 43 (27±71) Weight (kg) 66 (42±74) 62 (47±95) Height (cm) 169 (150±175) 168 (157±176) Duration of surgery (min) 185 (130±280) 180 (127±280) Peroperative propofol (mg) 2440 (1427±3266) 2123 (1070±3300) Peroperative alfentanil (mg) 1 (1±3) 1 (1±3) Peroperative haemorrhage (ml) 600 (200±1200) 500 (200±1900) Peroperative SAG M (ml) 0 (0±600) 0 (0±600) Peroperative sodium chloride (ml) 2000 (1000±3000) 2000 (1200±3600) Peroperative 6% HAES (ml) 500 (0±2000) 1000 (0±2000) Peroperative ephedrine (mg) 0 (0±55) 0 (0±45)

h h h h h h h h

P

Bupi

Bupi/morph

2 3 2 4 10 5 3 4

11 3 0 0 2 4 7 5

(1±15) (0±8) (0±13) (0±7) (2±31) (1±18) (0±8) (2±16)

(3±19) (0±10) (0±7) (0±3) (0±12) (0±11) (0±13) (0±20)

0.19 0.92 0.31 0.03 0.08 0.24 0.26 0.79

Table 3 VAS pain scores during cough [median (interquartiles)]

group; standard deviation of mean time to ®rst postoperative defaecation, 24 h. The sample size needed to demonstrate this difference was 15 in each group. Statistical analyses were performed using the Mann±Whitney rank sum test for unpaired data, the Wilcoxon signed rank sum test for paired data and Fisher's exact test for dichotomous data, where appropriate. A Cox regression analysis with time to ®rst postoperative ¯atus and defaecation as dependent variable was performed. Independent covariates chosen beforehand were treatment group, duration of operation, dose of supplementary morphine administered before the event, and age of the patient. If multiple testing was performed, signi®cant P values were corrected with a Bonferroni factor for multiple comparisons. P<0.05 was considered statistically signi®cant. Calculations were performed using SPSS 9.5 for Windows.

Postoperative time

Group Bupi

Bupi/morph

4 6 24 48 50 54 72 96

7 16 16 20 26 20 13 22

20 17 16 11 18 14 25 22

h h h h h h h h

(4±40) (8±28) (4±43) (4±38) (9±54) (6±45) (3±35) (4±35)

P

(12±44) (6±30) (5±43) (5±22) (8±28) (7±32) (13±35) (7±26)

0.17 0.94 0.95 0.36 0.50 0.63 0.27 0.72

Table 4 VAS pain scores during mobilization [median (interquartiles)] Postoperative time

Group Bupi

Bupi/morph

4 6 24 48 50 54 72 96

16 17 12 7 16 20 11 9

25 18 14 12 14 19 20 10

h h h h h h h h

(3±48) (7±46) (4±40) (2±19) (6±38) (6±32) (3±31) (4±18)

P

(12±39) (6±29) (2±29) (4±18) (10±28) (6±28) (9±35) (8±26)

0.31 0.84 0.97 0.48 0.96 0.79 0.37 0.23

Results We considered 53 consecutive patients for inclusion in the study. Thirteen were not included: ®ve used analgesics or had a history of alcohol abuse; three had contraindications to NSAIDs; one had preoperative chemotherapy; one did not want to participate; in one it was not possible to insert an epidural catheter; in one the preoperative epidural bolus dose produced a heterogeneous sensory block; and in one the epidural catheter became occluded during the operation. Thus, 40 patients were included in the study. Four of these were excluded during the study period, one because of change in surgical approach and three at 24 h (severe hypotension, accidental removal of the epidural catheter and withdrawal of consent because of severe itching). The patient with severe itching was in the bupi/morph group and the other three patients were in the bupi group. There were no signi®cant differences between groups with respect to patient or operative characteristics (Table 1). In particular, intraoperative hypotension was not a problem

and only nine and six patients in the bupi and the bupi/ morph groups respectively needed i.v. ephedrine. There were no signi®cant differences between groups for VAS pain scores at rest, during cough or during mobilization from the supine to the sitting position at any time (Tables 2±4). VAS pain scores increased signi®cantly in both groups at rest, during cough and during mobilization from 48 to 50 h after surgery, after the epidural catheters had been removed (P<0.03). Patients in the bupi group requested signi®cantly greater amounts of supplementary ketorolac and morphine from 24±48 h compared with the bupi/morph group (Tables 5 and 6). According to the Cox regression analysis, times to ®rst postoperative passage of ¯atus and defaecation were signi®cantly shorter in the bupi group than in the bupi/ morph group (¯atus, P=0.03; defaecation, P=0.02) (Fig. 1A and B). Times to ®rst postoperative ¯atus and defaecation

729

Jùrgensen et al. Table 5 Number of patients who requested ketorolac and number of doses administered. *P<0.05 after Bonferroni correction Postoperative period

0±24 h No. of No. of 24±48 h No. of No. of 48±72 h No. of No. of

Group

P

Bupi

Bupi/ morph

patients receiving ketorolac doses in patients receiving ketorolac

19/19 38

19/20 25

1.0 0.021*

patients receiving ketorolac doses in patients receiving ketorolac

15/17 22

4/18 8

0.001* 0.006*

patients receiving ketorolac doses in patients receiving ketorolac

14/17 27

15/18 32

1.0 0.53

Table 6 Number of patients who requested morphine and number of doses administered. *P<0.05 after Bonferroni correction Postoperative period

0±24 h No. of No. of 24±48 h No. of No. of 48±72 h No. of No. of

Group

P

Bupi

Bupi/ morph

patients receiving morphine doses in patients receiving morphine

18/19 43

18/20 28.5

0.13 0.07

patients receiving morphine doses in patients receiving morphine

9/17 12

2/19 2

0.03* 0.02*

patients receiving morphine doses in patients receiving morphine

11/17 24

7/19 11

0.18 0.15

Fig 1 Survival plots of times to ®rst ¯atus (A) and ®rst defaecation (B). Time was reduced signi®cantly in the bupi group (¯atus, P=0.03; defaecation, P=0.02; Cox regression analysis).

were signi®cantly in¯uenced by dose of supplementary morphine administered before the event. Duration of surgery and age of the patient did not have any in¯uence on times to ®rst postoperative ¯atus or defaecation, according to the Cox regression analysis. One patient in the bupi group suffered severe postoperative paralytic ileus. On day 5, she had a rectal tube inserted with some bene®t. Her ®rst postoperative spontaneous defaecation was 212 h after surgery. This patient was included in all analyses. There were no differences in the spread of sensory block or motor block at any assessment. Only three of 17 (18%) patients in the bupi group compared with six of 19 (32%) in the bupi/morph group and seven of 17 (41%) in the bupi group compared with eight of 19 (44%) in the bupi/morph group were able to walk without dif®culty at 24 and 48 h (no signi®cant difference). In 10 of 17 (59%) patients in the bupi group compared with 13 of 19 (68%) in the bupi/morph group and in six of 17 (35%) in the bupi group compared with 10 of 19 (53%) in the bupi/morph group, ability to walk could not be evaluated for reasons other than motor block (no signi®cant difference). In both groups, the primary reason for this was dizziness in approximately 70% of the patients 24 and 48 h after surgery. There was no difference between groups in mobilization time 24±48, 48±72 or 72±96 h after surgery. At 24±48 h, 11

and 16% of patients in the bupi group and bupi/morph groups respectively mobilized for more than 4 h. At 48±72 h the corresponding numbers were 29 and 21% and those at 72±96 h were 59 and 58%. The two groups had similar incidences of nausea and vomiting (Table 7). There were no signi®cant differences in the number of patients requesting antiemetics or the number of doses (Table 8). Nausea was mild in most patients; however, nine patients (®ve in the bupi group and four in the bupi/morph group) experienced severe nausea. Nine patients reported nausea of very short duration (minutes) during attempts to mobilize. The incidence of itching was signi®cantly different in the two groups. At 24 h, one patient in the bupi group compared with nine in the bupi/morph group suffered itching (P=0.008) and at 48 h the corresponding numbers were one and eight (P=0.02) (Table 7). Furthermore, 24 h after surgery one patient in the bupi/ morph group withdrew consent because of intolerable itching. All patients had an indwelling bladder catheter for 4 days. Therefore, no patients experienced urinary retention. There were no signi®cant differences between groups with respect to amount and time to ingestion of food and beverages. Patients ful®lled discharge criteria a median of 5

730

Epidural bupivacaine vs bupivacaine/morphine Table 7 Number of patients who experienced nausea and/or vomiting since last assessment, and number of patients suffering from itching. *P<0.05 after Bonferroni correction Postoperative period

0±24 h No. of No. of No. of 24±48 h No. of No. of No. of 48±72 h No. of No. of No. of

Group

P

Bupi

Bupi/ morph

patients experienced nausea patients who vomited patients suffering from itching

5/19 6/19 1/19

9/20 7/20 9/20

0.32 1.0 0.016*

patients experienced nausea patients who vomited patients suffering from itching

6/17 3/17 1/17

9/19 7/19 8/19

0.52 0.27 0.04*

patients experienced nausea patients who vomited patients suffering from itching

5/17 3/17 0/17

4/19 1/19 0/19

0.71 0.33 1.0

Table 8 Number of patients who requested antiemetics (ondansetron or metoclopramide) and number of doses administered Postoperative period

0±24 h No. of No. of 24±48 h No. of No. of 48±72 h No. of No. of

Group

P

Bupi

Bupi/ morph

patients receiving antiemetics doses in patients receiving antiemetics

8/19 9

11/20 16

0.52 0.26

patients receiving antiemetics doses in patients receiving antiemetics

6/17 8

8/19 20

0.74 0.45

patients receiving antiemetics doses in patients receiving antiemetics

4/17 4

5/19 13

1.0 0.68

(quartiles 4±7) days and 6 (5±7) days after operation in the bupi and bupi/morph groups respectively (not signi®cant).

Discussion Provided that the epidural catheter is placed at the correct intervertebral space, epidural analgesia with local anaesthetic is a powerful method of relieving postoperative pain.14 After abdominal surgery, the combination of epidural local anaesthesia and opioid has been shown to produce superior pain relief compared with the same dose of epidural local anaesthesia alone.1 12 15±18 Very few studies have focused on both postoperative pain relief and postoperative gastrointestinal paralysis. Improved pain relief may be at the expense of impaired gastrointestinal function.5 This study was designed to compare the effects of postoperative epidural bupivacaine 2 mg ml±1/morphine 50 mg ml±1 with the effects of the double dose and volume of epidural bupivacaine alone, on pain and gastrointestinal function. Because of the lack of dose±response data for

epidural bupivacaine, the dose administered to patients receiving epidural bupivacaine alone was chosen on the basis of a previous study conducted at our institution.19 Intra- and postoperative administration of opioids was restricted and standardized to reduce gastrointestinal paralysis and PONV. Thus, paracetamol was administered to all patients and ketorolac was the ®rst-choice escape analgesic. Morphine was administered after surgery only if ketorolac was insuf®cient. We showed no differences in VAS pain scores at rest, or during cough and ambulation between the study groups at any time, but a signi®cantly increased need for both ketorolac and morphine in patients receiving bupivacaine alone. This is a clear indication that the combination of epidural local anaesthetic and opioid provides superior analgesia compared with the double dose of epidural local anaesthetic alone. Regression analysis showed a reduction in time to restoration of gastrointestinal function in the bupi group compared with the bupi/morph group; there was a mean difference of 3 h to ®rst postoperative ¯atus and 18 h to ®rst postoperative defaecation. These results are in accordance with the ®ndings by Asantila and colleagues12 and demonstrate that either omitting a small dose of epidural morphine or an increased dose of bupivacaine may improve the restoration of gastrointestinal function. Furthermore, regression analysis indicated that prolongation of time to ®rst defaecation was associated with administration of supplementary systemic morphine. As in previous studies,20 administration of epidural morphine was associated with itching in almost 50% of patients and one patient withdrew from the study because of this side-effect. A high incidence of this relatively minor side-effect may be acceptable if the patients get clear bene®ts. In this study, however, the bene®t of the epidural combination of local anaesthetic and opioid was restricted to a reduced need for supplementary analgesics. Gastrointestinal function was impaired, itching was a signi®cant problem and no other differences in side-effects were demonstrated. However, we studied a small number of patients, so conclusions should be guarded. In summary, in patients undergoing major gynaecological surgery with free access to postoperative ketorolac and systemic morphine, epidural bupivacaine alone resulted in reduced gastrointestinal paralysis and itching, but similar pain relief and PONV, when compared with epidural bupivacaine and morphine. Under the conditions of this study, no effect on recovery or time to discharge from hospital was observed.

Acknowledgements The study was funded by a grant from Snedkermester Axel Wichmann og fru Else Wichmann's Fond. We thank research nurse Lene Henkel for much help and ¯exibility in assessing patients.

731

Jùrgensen et al.

References 1 Brodner G, Mertes N, van Aken H et al. Epidural analgesia with local anesthetics after abdominal surgery: earlier motor recovery with 0.2% ropivacaine than 0.175% bupivacaine. Anesth Analg 1999; 88: 128±33 2 Hovorka J, Korttila K, Erkola O. Nausea and vomiting after general anaesthesia with iso¯urane, en¯urane or fentanyl in combination with nitrous oxide and oxygen. Eur J Anaesthesiol 1988; 5: 177±82 3 Madej TH, Simpson KH. Comparison of the use of domperidone, droperidol and metoclopramide in the prevention of nausea and vomiting following major gynaecological surgery. Br J Anaesth 1986; 58: 884±7 4 Tsui SL, Ng KF, Wong LC, Tang GW, Pun TC, Yang JC. Prevention of postoperative nausea and vomiting in gynaecological laparotomies: a comparison of tropisetron and ondansetron. Anaesth Intens Care 1999; 27: 471±6 5 Jorgensen H, Wetterslev J, Moiniche S, Dahl JB. Epidural local anaesthetics versus opioid-based analgesic regimens on postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery (Cochrane Review). In: The Cochrane Library, Issue 3. Oxford: Update Software, 2001 6 Holte K, Kehlet H. Postoperative ileus: a preventable event. Br J Surg 2000; 87: 1480±93 7 Livingston EH, Passaro EP Jr. Postoperative ileus. Digestion Dig Sci 1990; 35: 121±32 8 Glise H, Abrahamsson H. Re¯ex inhibition of gastric motility pathophysiological aspects. Scand J Gastroenterol 1984; 89 Suppl: 77±82 9 Thoren T, Tanghoj H, Mattwil M, Jarnerot G. Epidural morphine delays gastric emptying and small intestinal transit in volunteers. Acta Anaesthesiol Scand 1989; 33: 174±80 10 Thoren T, Wattwil M. Effects on gastric emptying of thoracic epidural analgesia with morphine or bupivacaine. Anesth Analg 1988; 67: 687±94 11 Thorn SE, Wattwil M, Kallander A. Effects of epidural morphine

732

12

13 14

15

16

17 18

19

20

and epidural bupivacaine on gastroduodenal motility during the fasted state and after food intake. Acta Anaesthesiol Scand 1994; 38: 57±62 Asantila R, Eklund P, Rosenberg PH. Continuous epidural infusion of bupivacaine and morphine for postoperative analgesia after hysterectomy. Acta Anaesthesiol Scand 1991; 35: 513±7 Liu SS, Carpenter RL, Mackey DC et al. Effects of perioperative analgesic technique on rate of recovery after colon surgery. Anesthesiology 1995; 83: 757±65 Scott DA, Beilby DS, McClymont C. Postoperative analgesia using epidural infusions of fentanyl with bupivacaine. A prospective analysis of 1,014 patients. Anesthesiology 1995; 83: 727±37 Cooper DW, Ryall DM, McHardy FE, Lindsay SL, Eldabe SS. Patient-controlled extradural analgesia with bupivacaine, fentanyl, or a mixture of both, after Caesarean section. Br J Anaesth 1996; 76: 611±5 George KA, Chisakuta AM, Gamble JA, Browne GA. Thoracic epidural infusion for postoperative pain relief following abdominal aortic surgery: bupivacaine, fentanyl or a mixture of both? Anaesthesia 1992; 47: 388±94 Lee A, Simpson D, Whit®eld A, Scott DB. Postoperative analgesia by continuous extradural infusion of bupivacaine and diamorphine. Br J Anaesth 1988; 60: 845±50 Scott DA, Blake D, Buckland M, Etches R, Halliwell R, Marsland C et al. A comparison of epidural ropivacaine infusion alone and in combination with 1, 2, and 4 mg/mL fentanyl for seventy-two hours of postoperative analgesia after major abdominal surgery. Anesth Analg 1999; 88: 857±64 Jorgensen H, Fomsgaard JS, Dirks J, Wetterslev J, Dahl JB. Effect of continuous epidural 0.2% ropivacaine vs 0.2% bupivacaine on postoperative pain, motor block and gastrointestinal function after abdominal hysterectomy. Br J Anaesth 2000; 84: 144±50 Gedney JA, Liu EH. Side-effects of epidural infusions of opioid bupivacaine mixtures. Anaesthesia 1998; 53: 1148±55