Patient-controlled analgesia: epidural fentanyl and i.v. morphine compared after Caesarean section

Transcription

1 British Journal of Anaesthesia 82 (3): (1999) Patient-controlled analgesia: epidural fentanyl and i.v. morphine compared after Caesarean section D. W. Cooper, U. Saleh, M. Taylor, S. Whyte, D. Ryall, M. S. Kokri, W. R. Desira, H. Day and E. McArthur South Cleveland Hospital, Marton Road, Middlesbrough, Cleveland TS4 3BW, UK We have compared patient-controlled epidural fentanyl (PCEF) and patient-controlled i.v. morphine (PCIM) after Caesarean section in 84 patients, in a randomized, double-blind study. All patients had an epidural and an i.v. patient-controlled analgesia (PCA) device, one of which delivered normal saline. Group PCEF received epidural fentanyl 20 µg with a 10-min lockout. Group PCIM received i.v. morphine 1 mg with a 5-min lockout. PCA use was lower for PCEF patients (P ). The highest pain score recorded at rest for PCEF patients was median 20 (interquartile range 10 33) mm compared with 32 (14 52) mm for PCIM patients (P 0.02). The highest pain score recorded on coughing was 31 (21 41) mm with PCEF compared with 56 (30 71) mm for PCIM (P 0.001). There was less nausea (P 0.02) and drowsiness (P ) with PCEF. There was no difference in the overall incidence and severity of pruritus (P 0.77). However, pruritus started earlier with PCEF. Br J Anaesth 1999; 82: Keywords: analgesia, patient-controlled; analgesics opioid, morphine; analgesics opioid, fentanyl; analgesic techniques, epidural; anaesthesia, obstetric Accepted for publication: September 28, 1998 After Caesarean section, epidural fentanyl produces similar analgesia to epidural morphine, but with less nausea and pruritus. 1 3 In contrast with epidural morphine, the relatively rapid onset and short to medium duration of action of epidural fentanyl makes it suitable for administration by patient-controlled analgesia (PCA). Provided bolus doses are small, and the lockout interval is sufficiently long, overdose is unlikely to occur with PCA. 4 Patient-controlled epidural fentanyl (PCEF) has been used for analgesia after Caesarean section PCEF produced similar analgesia to epidural morphine, 3 and to patientcontrolled epidural analgesia (PCEA) using pethidine. 7 8 Compared with PCEA using 0.1% bupivacaine, PCEF was found to provide similar analgesia on coughing, but to be more effective at providing analgesia at rest, without producing lower limb weakness. 6 In the same study, PCEF was compared with PCEA using a mixture of epidural bupivacaine and fentanyl. With PCEF, the dose of fentanyl administered was higher, but analgesia was similar and there were fewer side effects. PCEF has been associated with a remarkably low incidence of nausea (5%) and vomiting (0%). 56 Despite these findings, PCEF is not used widely. This is partly because of the continuing debate as to whether or not epidural fentanyl works predominantly by a spinal or systemic action However, experience with PCEF after Caesarean section suggests that it is more effective than patient-controlled i.v. fentanyl. 58 Patient-controlled i.v. morphine (PCIM) is used widely to provide analgesia after Caesarean section. Pain scores with PCIM have been reported to be higher than with epidural morphine, but satisfaction with analgesia was similar, and there was less pruritus Unfortunately, nausea and sedation are frequent side effects of PCIM. The aim of this study was to assess if there are any clinical benefits from using PCEF compared with PCIM after Caesarean section. The PCEF bolus dose and lockout interval used were those which have been found previously to be effective. 56 PCIM dose and settings were those used routinely at this hospital. Patients and methods After obtaining approval from the Local Hospital Ethics Committee and informed consent, we studied 84 ASA I II patients undergoing morning elective Caesarean section under regional anaesthesia, in a randomized, double-blind study. Patients who had a contraindication to the use of diclofenac were excluded. Intrathecal and epidural anaesthesia were administered at L3 4, or at an adjacent space. Intrathecal injection was performed using a 26-gauge pencil-point needle passed through an epidural 16-gauge Tuohy needle. Patients received intrathecal injection of 0.5% plain bupivacaine ml. Epidural fentanyl, 50 µg in 10 ml with normal British Journal of Anaesthesia

2 Epidural fentanyl or i.v. morphine? saline, was then given via the Tuohy needle. An epidural catheter was passed, leaving 4 5 cm in the epidural space. During the course of operation, a total of 10 ml of 0.5% bupivacaine were given in small increments via the epidural catheter. More epidural 0.5% bupivacaine was given, if required, to achieve a block above T6. I.v. alfentanil was used to treat breakthrough pain. At the end of operation, patients received rectal diclofenac 100 mg. In the recovery room, patients were attached to epidural and i.v. PCA devices. The epidural PCA device was programmed to allow a 5-ml bolus with a 10-min lockout interval, and the i.v. PCA device to allow a 0.5-ml bolus with a 5-min lockout. Group PCEF received epidural fentanyl 4 µg ml 1 from one PCA device and i.v. saline from the other. Group PCIM received epidural saline from one PCA device and i.v. morphine 2 mg ml 1 from the other. Patient grouping was decided by opening a blank envelope from a set of 84 shuffled envelopes containing the code for group PCEF (n 42) or for group PCIM (n 42). The investigator who set up the PCA devices was not involved with subsequent patient management or assessment. Patients, nursing and medical staff were unaware of patient grouping. The study lasted until 08:00 on the first postoperative morning. Patients were kept in the recovery area for 4 h after delivery and then transferred to a postnatal ward. They were told that only one PCA device contained analgesic solution so that both PCA buttons had to be pressed, up to every 5 min, to receive analgesia. An oral preparation of paracetamol with codeine was given as required for breakthrough pain. No other analgesics were given. If analgesia was judged to be inadequate by the patient herself, midwife or medical staff, the study was stopped and alternative analgesia given. Volumes of epidural and i.v. solution used were recorded at 4 and 8 h after delivery, at 23:00 and at 08:00 on the first postoperative morning. At the same times, visual analogue scores for pain (VAS) at rest and on coughing (100-mm scale: 0 no pain, 100 mm worst pain imaginable) were recorded. The presence or absence of nausea, vomiting, pruritus and drowsiness was recorded 15 min after operation (recovery), and at the following times: recovery 4 h, 4 8 h, 8 h 23:00 and 23:00 08:00 (nausea score: 0 none, 1 nausea without vomiting, 2 vomiting; pruritus and drowsiness scores: 0 none, 1 mild, 2 moderate, 3 severe). Drowsiness was not measured at night. Ventilatory frequency and sedation scores were recorded at regular intervals as part of routine monitoring in patients with PCA. A verbal rating score (VRS) for satisfaction with analgesia (0 excellent, 1 adequate, 2 fair, 3 poor) was recorded for those patients completing the study. VRS was recorded as 3 if the patient was withdrawn because of inadequate analgesia. A computer package (SPSS) was used for statistical Table 1 Patient data in group PCEF (patient-controlled epidural fentanyl) and group PCIM (patient-controlled i.v. morphine) (median (interquartile range)) Group PCEF Group PCIM n Age (yr) 31 (28 33) 29 (25 33) Height (cm) 156 ( ) 160 ( ) Weight (kg) 73 (62 79) 73 (64 85) Previous Caesarean section 63% 62% Time of delivery 11:00 (10:30 12:00) 12:00 (11:00 12:30) Tubal surgery 23% 7% Intrathecal bupivacaine (mg) 10 (10 10) 10 (10 10) Epidural fentanyl (µg) Epidural bupivacaine (mg) 50 (50 50) 50 (50 50) I.v. alfentanil given 6% 12% Rectal diclofenac given 100% 100% analysis. Results were compared using Wilcoxon signed rank and Mann Whitney tests, with P 0.05 as significant. Results We studied 84 patients, 42 in each group. Ten epidural catheters became dislodged in the postoperative period: seven in group PCEF and three in group PCIM. Data from the seven PCEF patients were excluded because it was unclear exactly when their epidural catheters had become dislodged. Exclusion of data from the seven PCEF patients did not significantly alter the results. For the three PCIM patients, only data obtained before the discovery of the dislodged epidural catheters were included. In addition, one PCEF and one PCIM patient withdrew between 8 h and 23:00 because of inadequate analgesia. Data were included before the premature cessation of the study for both patients. It was unclear if the catheter was placed correctly for the PCEF patient because it was not tested with local anaesthetic before removal. The groups were matched for age, height, weight, previous Caesarean section and operative analgesia (Table 1). There was a trend towards more PCEF patients also having tubal surgery at operation (P 0.051). The time of delivery was similar for both groups. To the nearest hour, the median 23:00 and 08:00 assessment times were 12 and 21 h after delivery, respectively. The overall volumes of epidural and i.v. PCA solution delivered per hour were lower for PCEF patients. The i.v. volume delivered per hour was median 0.45 (interquartile range (IQR) ) ml for group PCEF compared with 0.83 ( ) ml for group PCIM (P ; Mann Whitney). The epidural volume delivered per hour was 4.1 ( ) ml for group PCEF compared with 7.1 ( ) ml for group PCIM (P 0.001). The ratio of the median number of epidural to i.v. boluses delivered per hour was 0.91 for group PCEF and 0.86 for group PCIM. For PCEF patients, the overall dose of fentanyl used per hour was median 16 (IQR 11 29) µg. For PCIM patients, the overall dose of morphine used per hour was median 1.7 (IQR ) mg. The highest overall dose of fentanyl 367

3 Cooper et al. Table 3 Incidence (%) of postoperative nausea, vomiting, pruritus and drowsiness for each time, and overall incidence after commencement of PCA in group PCEF (patient-controlled epidural fentanyl) and group PCIM (patientcontrolled i.v. morphine). Nausea and vomiting, pruritus and drowsiness scores were compared: *P 0.05, **P 0.01, ***P (Mann Whitney). Rec. Recovery Rec. Rec. 4 h 4 8 h 8 12 h h Overall Fig 1 Visual analogue scores (VAS) for pain at rest (median (interquartile range)) for group PCEF (patient-controlled epidural fentanyl) and group PCIM (patient-controlled i.v. morphine). **P 0.01 (Mann Whitney). Fig 2 Visual analogue scores (VAS) for pain on coughing (median (interquartile range)) for group PCEF (patient-controlled epidural fentanyl) and group PCIM (patient-controlled i.v. morphine). *P 0.05, ***P (Mann Whitney). Table 2 Postoperative drug requirements in group PCEF (patient-controlled epidural fentanyl) and group PCIM (patient-controlled i.v. morphine) (median (interquartile range)) 0 4 h 4 8 h 8 12 h h PCEF (µg h 1 ) 6 (0 10) 16 (9 29) 20 (10 25) 18 (11 26) PCIM (mg h 1 ) 1.0 ( ) 2.0 ( ) 1.9 ( ) 1.9 ( ) used was 49 µg h 1. Within each group, from 4 h onwards, the dose of drug administered did not alter significantly from one assessment period to the next (Wilcoxon signed rank test) (Table 2). Pain scores at rest were lower for PCEF than for PCIM patients at 4 and 8 h after delivery (Fig. 1). Pain scores on coughing were lower with PCEF at 4, 8 and 21 h after delivery (Fig. 2). The highest pain score recorded at rest for PCEF patients was median 20 (IQR 10 33) mm compared with 32 (14 52) mm for PCIM patients (P 0.02). No PCEF patient had a pain score greater than 50 mm at rest compared with 31% of PCIM patients. The highest pain score recorded on coughing was 31 (21 41) mm for PCEF compared with 56 (30 71) mm for PCIM (P 0.001). Eleven percent of PCEF patients had a pain score greater than 50 mm on coughing compared with 51% of PCIM patients. Within both groups, pain scores at rest and on coughing increased significantly from recovery to 4 h, and from 4 to PCEF/PCIM Nausea 3/2 6/2 6/20 3/8 9/16 9/31* Vomiting 0/0 0/0 3/5 0/3 0/5 3/12 * Pruritus 7/15 31/39 54/24* 55/26* 54/51 63/62 Drowsiness 29/23 31/36 27/54** 27/49* 14/58*** 40/76*** 8h(P 0.01; Wilcoxon signed rank test). In group PCEF, there were no further significant changes in pain scores from one assessment time to the next (P 0.8). In group PCIM, pain scores at rest and on coughing decreased significantly from 8 to 12 h (P 0.03), and then increased significantly from 12 to 21 h (P 0.02). Table 3 shows the incidence of side effects with time, and the overall incidence after commencement of PCA. Overall, there was less nausea, vomiting and drowsiness with PCEF: 40% of PCEF patients were drowsy (79% mild, 21% moderate) compared with 76% of PCIM patients (50% mild, 44% moderate, 6% severe) (P ). The difference in drowsiness was present from 4 8 h onwards, but was most marked on the first postoperative morning. Overall, the incidence and severity of pruritus were similar: 63% of PCEF patients had pruritus (55% mild, 36% moderate, 10% severe) compared with 62% of PCIM patients (65% mild, 19% moderate, 15% severe) (P 0.77). However, pruritus was more frequent with PCEF at 4 12 h. Satisfaction with analgesia was similar in both groups: 42% of PCEF patients rated analgesia as excellent, 45% good, 9% fair and 3% poor. This compared with 35% of PCIM patients who rated analgesia as excellent, 49% good, 14% fair and 3% poor (P 0.5; Mann Whitney). No patient had respiratory depression (ventilatory frequency 10 bpm) or excessive sedation (somnolent, difficult to arouse). Discussion More than 80% of patients in each group rated postoperative analgesia as good or excellent. However, there were advantages in using PCEF compared with PCIM. PCA use, as measured by the median volume of i.v. solution delivered per hour, was 46% lower with PCEF. Pain scores were lower with PCEF. At rest, no PCEF patient had a pain score greater than 50 mm compared with 31% of PCIM patients. On coughing, only 11% of PCEF patients had a pain score greater than 50 mm compared with 51% of PCIM patients. PCEF patients had less nausea and drowsiness than PCIM patients. The difference in drowsiness was most marked on the first postoperative morning when only 14% of PCEF patients reported feeling drowsy compared with 58% of PCIM patients. 368

4 Epidural fentanyl or i.v. morphine? Our operative anaesthetic technique was chosen to try and minimize postoperative opioid requirements and side effects. Intrathecal fentanyl, given at operation, can increase postoperative morphine requirements, perhaps by inducing acute spinal opioid tolerance. 14 Epidural fentanyl has not been associated with increased postoperative opioid requirements. 15 Therefore, intrathecal bupivacaine was supplemented with epidural rather than with intrathecal fentanyl. Epidural fentanyl was administered in a total volume of 10 ml because this has been associated with more rapid onset and prolongation of analgesia. 16 Epidural bupivacaine given at operation can reduce opioid requirements in the early postoperative period. 17 Epidural bupivacaine 50 mg was therefore given routinely during the course of operation. Diclofenac was given at the end of operation to further contribute to reduced postoperative opioid requirements The results of our study contrast with a previous study which compared i.v. fentanyl PCA with i.v. morphine PCA after general anaesthesia for Caesarean section. 22 In that study, initial PCA settings allowed for a 25-µg bolus of i.v. fentanyl or for a 1-mg bolus of i.v. morphine, both with a 10-min lockout interval. To produce satisfactory analgesia, 72% of i.v. fentanyl patients required extra boluses or PCA setting adjustments compared with only 21% of i.v. morphine patients. This resulted in average hourly i.v. fentanyl requirements being 3.8 times greater than for epidural fentanyl in our study. I.v. morphine requirements were 0.85 of those in our study. Results from these two studies suggest PCA with epidural fentanyl is more effective than with i.v. fentanyl. This is in agreement with the results of previous studies comparing epidural fentanyl PCA with i.v. fentanyl PCA after Caesarean section. 58 The most likely explanation is that epidural fentanyl has a predominantly spinal action when it is administered in the minimal effective analgesic dose. Morphine requirements were similar to those from a previous study at this hospital when PCIM was used after comparable combined spinal epidural anaesthesia for Caesarean section. 14 Rectal diclofenac was given to 93% of patients in that study. In this study, epidural fentanyl requirements were 35% lower than found previously using an identical PCEF regimen after epidural anaesthesia for Caesarean section at this hospital. 56 This was probably because of the opioid sparing effect of diclofenac, which was not used in the previous PCEF studies. It is also possible that some intrathecal spread of fentanyl occurred through the small hole made in the dura with combined spinal epidural anaesthesia. Such a theory has been proposed previously to explain the greater analgesic efficacy of epidural diamorphine after combined spinal epidural than after epidural anaesthesia. 23 Mean hourly PCA use was constant within each group from 4 h after delivery onwards, but pain scores were not. With PCEF, pain scores increased until 8 h after delivery, remaining constant thereafter. With PCIM, pain scores were higher than with PCEF at 4 and 8 h, but by 12 h they had decreased to similar levels. Surprisingly, from 12 to 21 h, pain scores increased with PCIM but remained constant with PCEF. This resulted in higher pain scores on coughing at 21 h for PCIM than for PCEF patients. Acute tolerance to opioids can develop. There is evidence that tolerance develops earlier with less potent opioids, such as morphine, than with more potent opioids, such as fentanyl. 24 The apparent decrease in efficacy of PCIM from 12 to 21 h, but not of PCEF, could be explained by the development of differential tolerance. Sedation has been reported previously to be less severe with epidural fentanyl than with i.v. morphine. 25 In our study there was less drowsiness and nausea with PCEF than with PCIM. Equi-analgesic systemic doses of fentanyl and morphine appeared to produce a similar incidence of central side effects. 22 The most likely reason for a difference in central side effects between epidural fentanyl and i.v. morphine is greater spinal selectivity with epidural fentanyl. Intrathecal and epidural fentanyl used to supplement regional anaesthesia for Caesarean section have both been associated with a reduction in nausea This may have been because of more effective visceral afferent inhibition with fentanyl. More effective analgesia with PCEF than with PCIM may therefore have contributed to the reduced nausea. Epidural fentanyl produces less pruritus than epidural morphine after Caesarean section. 1 3 It is generally considered that epidural opioids produce more pruritus than i.v. opioids. However, our results suggest that when epidural fentanyl and i.v. morphine are administered in minimal effective analgesic doses, the overall incidence and severity of pruritus is similar, although pruritus develops earlier with epidural fentanyl. Epidural fentanyl, administered either by infusion or by PCA to older patient populations than in our study, has been used safely in a ward environment In our study, PCEF settings could potentially have allowed more than 100 µg of fentanyl to be delivered per hour. However, the median dose of fentanyl administered was only 16 µg h 1 and the highest dose administered was 49 µgh 1. Therefore, it is possible that after Caesarean section, the margin of safety for PCEF could be increased further by the addition of a 50 µg h 1 dose limit without adversely affecting analgesia. Acknowledgements We thank Graseby Medical and Abbott Laboratories for the loan of PCA devices. References 1 Fischer RL, Lubenow RT, Liceaga A, McCarthy RJ, Ivankovich AD. Comparison of continuous epidural infusion of fentanyl bupivacaine and morphine bupivacaine in management of postoperative pain. Anesth Analg 1988; 67:

5 Cooper et al. 2 White MJ, Berghausen EJ, Dumont SW, et al. Side effects during continuous epidural infusion of morphine and fentanyl. Can J Anaesth 1992; 39: Yu PYH, Gambling DR. A comparative study of patient-controlled epidural fentanyl and single dose epidural morphine for post- Caesarean analgesia. Can J Anaesth 1993; 40: White PF. Use of patient-controlled analgesia for management of acute pain. JAMA 1988; 259: Cooper DW, Ryall DM, Desira WR. Extradural fentanyl for postoperative analgesia: predominant spinal or systemic action? Br J Anaesth 1995; 74: 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: Goh JL, Evans SF, Pavy TJG. Patient-controlled epidural analgesia following Caesarean delivery: A comparison of pethidine and fentanyl. Anaesth Intensive Care 1996; 24: Ngan Kee WD, Lam KK, Chen PP, Gin T. Comparison of patientcontrolled epidural analgesia with patient-controlled intravenous analgesia using pethidine or fentanyl. Anaesth Intensive Care 1997; 25: Cheam EWS, Morgan M. The superiority of epidural opioids for postoperative analgesia fact or fallacy? Anaesthesia 1994; 49: Chrubasik S, Chrubasik J. Selection of the optimum opioid for extradural administration in the treatment of postoperative pain. Br J Anaesth 1995; 74: de Leon-Casasola OA, Lema MJ. Postoperative epidural opioid analgesia: What are the choices? Anesth Analg 1996; 83: Eisenach JC, Grice SC, Dewan DM. Patient-controlled analgesia following Cesarean section: A comparison with epidural and intramuscular narcotics. Anesthesiology 1988; 68: Harrison DM, Sinatra R, Morgese L, Chung JH. Epidural narcotic and patient-controlled analgesia for post-cesarean section pain relief. Anesthesiology 1988; 68: Cooper DW, Lindsay SL, Ryall DM, Kokri MS, Eldabe SS, Lear GA. Does intrathecal fentanyl produce acute cross-tolerance to i.v. morphine. Br J Anaesth 1997; 78: Naulty JS, Datta S, Ostheimer GW, Johnson MD, Burger GA. Epidural fentanyl for postcesarean delivery pain management. Anesthesiology 1985; 63: Birnbach DJ, Johnson MD, Arcario T, Datta S, Naulty JS, Ostheimer GW. Effect of diluent volume on analgesia produced by epidural fentanyl. Anesth Analg 1989; 68: Lam FY, Broome IJ, Matthews PJ. A comparison of postoperative analgesia following spinal or epidural anaesthesia for Caesarean section. Anaesthesia 1994; 49: Bush DJ, Lyons G, Macdonald R. Diclofenac for analgesia after Caesarean section. Anaesthesia 1992; 47: Sun HL, Wu CC, Lin MS, Chang CF, Mok MS. Combination of low-dose epidural morphine and intramuscular diclofenac sodium in postcesarean analgesia. Anesth Analg 1992; 75: Luthman J, Kay NH, White JB. The morphine sparing effect of diclofenac sodium following Caesarean section under spinal anaesthesia. Int J Obstet Anaesth 1994; 3: Dennis AR, Leeson-Payne CG, Hobbs GJ. Analgesia after Caesarean section. The use of rectal diclofenac as an adjunct to spinal morphine. Anaesthesia 1995; 50: Howell PR, Gambling DR, Pavy T, McMorland G, Douglas MJ. Patient-controlled analgesia following Caesarean section under general anaesthesia: a comparison of fentanyl with morphine. Can J Anaesth 1995; 42: Norman BJ, Yentis S. Analgesia produced by epidural diamorphine is better following Caesarean section under spinal anaesthesia than under epidural anaesthesia. Int J Obstet Anaesth 1998; 7: Gebhart GF. Some mechanistic insights into opioid tolerance. Anesthesiology 1990; 73: Benzon HT, Wong HY, Belavic AM, et al. A randomized doubleblind comparison of epidural fentanyl infusion versus patientcontrolled analgesia with morphine for postthoracotomy pain. Anesth Analg 1993; 76: Dahlgren G, Hulstrand C, Jakobsson J, Norman M, Eriksson EW, Martin H. Intrathecal sufentanil, fentanyl, or placebo added to bupivacaine for Cesarean section. Anesth Analg 1997; 85: Vincent RD, Chestnut DH, Choi WW, Ostman PLG, Bates JN. Does epidural fentanyl decrease the efficacy of epidural morphine after Cesarean delivery? Anesth Analg 1992; 74: Scott DA, Beilby DSN, McClymont C. Postoperative analgesia using epidural infusions of fentanyl with bupivacaine: A prospective analysis of 1,014 patients. Anesthesiology 1995; 83: Liu SS, Allen HW, Olsson GL. Patient-controlled epidural analgesia with bupivacaine and fentanyl on hospital wards: prospective experience with 1,030 surgical patients. Anesthesiology 1998; 88: