Y. Lim, 1 A. T. Sia 2 and C. E. Ocampo 3

Transcription

1 Anaesthesia, 2006, 61, pages doi: /j x Comparison of computer integrated patient controlled epidural analgesia vs. conventional patient controlled epidural analgesia for pain relief in labour Y. Lim, 1 A. T. Sia 2 and C. E. Ocampo 3 1 Associate Consultant, 2 Head of Obstetric Anaesthesia and Senior Consultant, 3 Clinical Fellow, Department of Women s Anaesthesia, KK Women s and Children s Hospital, Singapore Summary Computer integrated-patient controlled epidural analgesia (CI-PCEA) is a novel drug delivery system. It automatically titrates the background infusion rate based on the individual parturient s need. In this randomised trial, we compared the local anaesthetic consumption by parturients using CI-PCEA with demand only patient controlled epidural analgesia (PCEA) for labour analgesia. We recruited 40 parturients after approval by the ethics committee. Group PCEA (n = 20) received demand only PCEA. Group CI-PCEA (n = 20) received a similar PCEA regimen but the computer integration titrated the background infusion to 5, 10 or 15 ml.h )1 if the patient required one, two or three demand boluses, respectively, in the previous hour. The background infusion decreased by 5 ml.h )1 if there was no demand bolus in the previous hour. The sample size was calculated to show equivalence in local anaesthetic used. The time weighted consumption of local anaesthetic was similar in both groups (mean difference 0.7 mg.h )1, 95% confidence interval [CI: )2.5, 1.1]; p = 0.425). The CI-PCEA group had higher maternal satisfaction scores: mean (SD) 93 (7) vs. 86 (11), p = CI-PCEA does not increase the use of local anaesthetic when compared with demand only PCEA but does increase patient satisfaction.... Correspondence to: Yvonne Lim yvel6@hotmail.com Accepted: 9 December 2005 Patient controlled epidural analgesia (PCEA) has been established as a safe and effective method of maintaining epidural analgesia for patients in labour [1]. Its advantages over conventional continuous epidural infusion techniques include patient autonomy in determining the level of pain relief, reduced local anaesthetic consumption, reduced motor block, improved pain scores, decreased anaesthetic workload, and the potential for improved maternal satisfaction [2]. However, the optimal PCEA regimen is still a subject of debate [2]. Initial studies seemed to favour regimens without a background infusion. These studies demonstrated that a background infusion with PCEA during labour leads to a greater consumption of anaesthetic solution without improving comfort and satisfaction of parturients or decreasing breakthrough pain [3, 4]. Recently, several studies have demonstrated the beneficial effects of having a background infusion. A PCEA regimen with a background infusion appeared to be more effective in lowering pain scores of parturients than demand only PCEA and the amount of local anaesthetic used did not appear to be increased [5, 6]. In reality, the need for more analgesics, including the requirement for background infusion, would be influenced by the stage and progress of labour [7, 8]. To address this problem, our centre has developed a novel drug delivery system (Fig. 1). Computer integrated-patient controlled epidural analgesia (CI-PCEA) is programmed to analyse the parturient s local anaesthetic requirement in the last 1 h and adjust the background infusion rate accordingly (Fig. 2). In this study, we compared CI-PCEA with demand only PCEA for the maintenance of labour epidural analgesia following a combined spinal epidural (CSE) technique. Our primary outcome was the time-weighted, hourly consumption of local anaesthetic used for the duration of labour epidural analgesia. We also assessed the incidence of breakthrough pain, duration of labour analgesia, side-effects and maternal satisfaction. Journal compilation Ó 2006 The Association of Anaesthetists of Great Britain and Ireland 339

2 Y. Lim et al. Æ Computer integrated-patient controlled epidural analgesia Anaesthesia, 2006, 61, pages Two-way communications between the pump and the laptop computer using RS232 serial ports CI-PCEA program source code compiled in Borland Pascal language on Windows Operating Systems Modified Infusion Pump (IVAC P700, Alaris, UK) Remote Control Laptop (IBM Thinkpad) with built-in program for computer integrated patient controlled epidural analgesia Figure 1 Computer integrated-patient controlled epidural analgesia. CSE: intrathecal fentanyl 15 µg + ropivacaine 2 mg + epidural 1.5% lidocaine 2 ml Method No infusion Demand dose: 5 ml *, lockout 10 min Change infusion to 5 ml.h 1 Demand within 1 h Change infusion to 10 ml.h 1 Demand within 1 h Change infusion to 15 ml.h 1 Demand within 1 h *0.1% ropivacaine + fentanyl 2 µg.ml 1 No demand in 1 h No demand in 1 h No demand in 1 h Stop infusion and activate alarm Figure 2 Schematic representation of CI-PCEA algorithm. With the approval of the hospital research ethics committee and informed written consent, we recruited 40 ASA I, nulliparous labouring parturients at term who had requested neuraxial analgesia into this randomised, double blinded, controlled trial. We included parturients in established labour with cervical dilation = 5 cm and with baseline pain scores 5 (on a 0 10 visual analogue scale (VAS): 0 = no pain, 10 = worst pain imaginable). We excluded parturients who had received parenteral opioids < 2 h earlier or had baseline pain scores < 5 obtained before the CSE. Parturients with obstetric complications, such as pre-eclampsia, multiple pregnancies, macrosomia, diabetes and malpresentation, were also excluded from our study. Each parturient received a preload of intravenous Ringer s lactate solution 500 ml for hydration. Baseline pain scores and systolic blood pressure (SBP), which was measured non-invasively on the right brachial artery (Dinamap, Critikon, FL) with the parturient supine and with left uterine displacement, were obtained before the CSE. The use of cervical prostaglandin E2, intravenous oxytocin and the degree of cervical dilation prior to the CSE were also recorded. Every parturient received a CSE in the left lateral decubitus position. After the epidural space was located with a 17-gauge Weiss needle using loss of resistance to < 2 ml of air, dural puncture was performed by passing a 27-gauge pencil-point spinal needle through the epidural needle (Espocan, B. Braun, Melsungen, Germany). After ensuring a free flow of cerebrospinal fluid (CSF), ropivacaine 2 mg and fentanyl 15 lg diluted with normal saline to a total volume of 2 ml was injected intrathecally over 15 s with the orifice of the spinal needle facing in the cephalad direction. A multi-orifice epidural catheter was then inserted 3 4 cm into the epidural space. The catheter was tested with 3 ml of 1.5% lidocaine to exclude intrathecal (IT) placement. If there was significant motor block (inability to flex the knees) or a reduction of > 20% in SBP, the patient was withdrawn from the study because of suspected IT catheter placement. Patients who had blood or CSF aspirated from the catheter were also withdrawn from the study. We randomly allocated the parturients using a sealed opaque envelope to two groups. Group PCEA received a PCEA regimen with no background infusion, an epidural bolus of 5 ml ropivacaine 0.1% with fentanyl 2 lg.ml )1 on demand, a lock-out period of 15 min and a maximum dose of 20 ml. h )1 (Rythmic TM Pump Micrel Medical Devices S.A., Pallini, Greece). Group CI-PCEA received a similar PCEA regimen using the same local anaesthetic and opioid solution. The 340 Journal compilation Ó 2006 The Association of Anaesthetists of Great Britain and Ireland

3 Anaesthesia, 2006, 61, pages Y. Lim et al. Æ Computer integrated-patient controlled epidural analgesia starting background infusion rate was zero but computer integration allowed the background infusion to increase by 5 ml.h )1 if the parturient required one demand bolus in the previous hour. If the patient required two demand boluses in the previous hour, the background infusion increased to 10 ml.h )1. The maximum background infusion rate was limited to 15 ml.h )1 and further demand boluses activated an alarm to alert the attending anaesthetist to review the patient. Likewise, if the patient did not receive a demand bolus in the previous hour, the background infusion decreased by 5 ml.h )1 (Fig. 2). The patients were blinded to their group allocation and were not told the PCEA regimen to which they were assigned. Both groups were given a handheld device and instructed to self-administer an epidural bolus dose by clicking a button. They were counselled to activate an epidural bolus when they experienced mild to moderate pain, and before the pain intensity became severe. Parturients who did not obtain satisfactory pain relief (VAS < 3) 15 min after the CSE were deemed to have a failed block. Rescue medication was then delivered via the epidural catheter and the patient was removed from the study. An anaesthetist who was not involved in performing the block, collected the following data at 5, 10, 15 and 30 min after the CSE technique and at 2, 4, 8, 10 and 12 h for those who had not delivered: Systolic blood pressure (SBP), maternal heart rate (HR) and fetal heart rate (FHR); Pain scores using the 0 10 VAS (0 = no pain, 10 = worst pain imaginable); Sensory level (tested with an ice cube); Side-effects: the presence or absence of lower limb motor block, pruritus, shivering, hypotension, nausea, vomiting and fetal bradycardia. Breakthrough pain was defined as failure of the PCEA or CI-PCEA to provide adequate pain relief, necessitating a review by the attending anaesthetist prior to delivery. The anaesthetist assessed the pain scores and 0.2% ropivacaine 5 ml bolus was given epidurally as rescue analgesia. If VAS remained above 3 despite administering a total of up to 20 ml of 0.2% ropivacaine epidurally (in 5 ml aliquots over 30 min), the catheter was labelled as ineffective and the subject excluded from analysis. After the breakthrough pain was relieved with additional analgesia, parturients in Group CI-PCEA were converted to the default regimen, i.e. conventional continuous epidural infusion: 0.1% ropivacaine and fentanyl 2 lg.ml )1 at a rate of 10 ml.h )1. In our institution, the use of continuous epidural infusion to maintain epidural analgesia has been the usual practice and the hospital ethics committee stipulated that when the patients randomly assigned to CI-PCEA encountered breakthrough pain, a conversion to the default regimen (i.e. continuous epidural infusion) was necessary, in keeping with the standard clinical practice of the institution. This also potentially made any comparison of the purported drug sparing effect of CI-PCEA with PCEA more conservative and reduced the risk of false positives. Fetal heart rate (from a continuous external cardiotocogram) was assessed by the attending obstetrician who was blinded to the drugs received by the parturients. The pain scores, cervical dilation and use of oxytocin at time END were also recorded. The time of delivery, mode of delivery, neonate Apgar scores, and overall satisfaction with neuraxial analgesia were assessed and documented within 2 h of delivery on a scale (0 = very dissatisfied to 100 = extremely satisfied). In previous studies, background infusion with PCEA during labour has been shown to increase the consumption of anaesthetic solution. CI-PCEA has a variable background infusion depending on the parturients need and we felt this would not increase consumption of local anaesthetic significantly. The null hypothesis in this equivalence trial states that the hourly consumption of anaesthetic solution in demand-only PCEA is less than that in CI-PCEA. The aim of the trial was to demonstrate equivalence in the hourly consumption of anaesthetic solution in both groups. CI-PCEA was considered to be at least equivalent to demand only PCEA if the 95% confidence interval of the difference in local anaesthetic consumption did not exceed 30% of the absolute hourly infusion rate. A difference of a lesser magnitude would probably not be of any clinical significance, especially when a lower concentration of local anaesthetic and opioids was used. We assumed the consumption of local anaesthetic with demand only PCEA to be 7 mg.h )1 [9]. Accordingly, we calculated that a sample size of 20 patients per treatment group was needed to reject the null hypothesis (of nonequivalence) with a power of 80% at the significance level of 5% if the specified alternative hypothesis (of equivalence) were true. For the primary end point, a 95% CI and p-value (testing for equivalence) are given. Analysis of dichotomous data (incidence of breakthrough pain and side-effects) was performed using Chi-squared tests. The Student t-test and Mann Whitney test were employed to analyse parametric and non-parametric data, respectively. All data and statistical analyses were managed with SPSS version 9 (Chicago, IL). Results Both groups were similar in terms of age, weight and height. There were also no differences in the preblock Journal compilation Ó 2006 The Association of Anaesthetists of Great Britain and Ireland 341

4 Y. Lim et al. Æ Computer integrated-patient controlled epidural analgesia Anaesthesia, 2006, 61, pages CI-PCEA group (n = 20) PCEA group (n = 20) p-value Weight; kg 66.7 (9.9) 65.6 (9.8) Height; cm (6.4) (5.8) Age; year 27.3 (4.7) 29.3 (3.5) Systolic blood pressure; mmhg 120 (11) 116 (15) Maternal heart rate; min )1 86 (23) 85 (21) Fetal heart rate; min )1 142 (9) 139 (17) Pre-block pain score; VAS 8.1 (1.8) 7.4 (1.6) Use of oxytocin (before induction 5 (25%) 3 (15%) of epidural analgesia) Use of prostaglandin E2 for 12 (60%) 11 (55%) induction of labour Pre-block cervical dilatation; cm 3 [2 5] 3 [2 5] Table 1 Preliminary demographic and obstetric data. Values are mean (SD), n (proportion) or median [range]. CI-PCEA, computer integrated-patient controlled epidural analgesia. PCEA, patient controlled epidural analgesia. cervical dilation score, pain scores, use of prostaglandin E2 for induction of labour, use of oxytocin for augmentation of labour, systolic blood pressure, maternal and fetal heart rate (Table 1). All parturients had VAS < 3 at 15 min after the CSE and there were no failed blocks or ineffective catheters. The time weighted consumption of local anaesthetic administered epidurally from the time of induction of the CSE to the time of delivery, including local anaesthetic used to treat breakthrough pain, was similar in the PCEA group (mean (SD) 7.2 (2.6) mg.h )1 ) and the CI-PCEA group (mean (SD) 7.9 (2.9) mg.h )1 ; time weight mean difference )0.7 mg.h )1 ): 95% CI [)2.5, 1.1], p = We also could not detect a difference in the total amount of local anaesthetic used, being 45.9 mg in the PCEA group and 39.1 mg in the CI-PCEA group with a mean difference 6.8 mg, 95% CI [)7.2, 20.8]; p = (Table 2). The CI-PCEA group also did not show a difference in incidence of breakthrough pain when compared with Group PCEA (15% vs. 35%, p = 0.144). Visual analogue scale, oxytocin usage and degree of cervical dilation at the time of breakthrough pain were similar in both groups (Table 3). Our analysis using the Kaplan-Meier technique, which took into account parturients who had delivered prior to the loss of analgesia as censored data, showed that the mean duration of analgesia in the CI-PCEA group was similar to the PCEA group (mean survival time min, 95% CI [593, 733] vs. 589 min, 95% CI [483, 696], p = 0.226) (Fig. 3). However, the CI-PCEA group had significantly higher maternal satisfaction scores with labour analgesia than the PCEA group: mean (SD) 93 (7) vs. 86 (11), p = None of the parturients developed significant motor block even though we did not attempt to ambulate them. There was no difference in the serial pain scores during the first 8 h postblock in both groups (Fig. 4). There was no difference in the sensory blockade level between the groups (Table 2). Both groups were similar in their CI-PCEA group (n = 20) PCEA group (n = 20) p-value Maximal dermatomal block to cold T6 [T9 T2] T6 [T8 T2] Max pain scores during 2.0 (3.0) 2.1 (2.1) maintenance of analgesia Lower limb motor block 1 (5%) 1 (5%) 1.0 (Bromage score > 0) Mode of delivery Normal vaginal delivery 18 (90%) 17 (85%) Instrumentation 0 (0%) 1 (5%) Caesarean section 2 (10%) 2 (10%) Duration of 2nd stage; min 67.2 (42.8) 79.2 (61.0) Duration of Labour; min (161.2) (188.4) Total LA used; mg 39.1 (19.7) 45.9 (23.8) Time-weighted LA used; mg.h )1 7.9 (2.9) 7.2 (2.6) Fetal birth weight; g 3216 (410) 3050 (518) Apgar scores at 5 min 9 [6 9] 9 [9 9] Table 2 Characteristics at of labour analgesia and obstetric outcome. Values are median [range], n (proportion) or mean (SD). LA, local anaesthetic. CI-PCEA, Computer integrated-patient controlled epidural analgesia. PCEA, Patient controlled epidural analgesia. 342 Journal compilation Ó 2006 The Association of Anaesthetists of Great Britain and Ireland

5 Anaesthesia, 2006, 61, pages Y. Lim et al. Æ Computer integrated-patient controlled epidural analgesia Table 3 Profile at breakthrough pain. Values are n (proportion), mean (SD) or median [range]. CI-PCEA group PCEA group p-value Breakthrough pain 3 (15%) 7 (35%) Pain score at breakthrough pain (VAS) 8.7 (1.5) 7.7 (1.8) Use of prostaglandin E2 at breakthrough pain 2 (67%) 7 (100%) Cervical dilation at breakthrough pain; cm 6 [3 6] 4 [3 5] No significant differences were found between the two groups. Values are tabulated from parturients who experienced breakthrough pain with the CI-PCEA group, n = 3 and the PCEA group, n = 7. CI-PCEA, computer integrated-patient controlled epidural analgesia. PCEA, patient controlled epidural analgesia. Proportion of undelivered parturients with effective analgesia Group CI PCEA Group CI-PCEA: delivery Group PCEA + Group PCEA: delivery Time after induction of labour epidural analgesia (min) Figure 3 Proportion of Computer integrated patient controlled epidural analgesia (CI-PCEA) and Patient controlled epidural analgesia (PCEA) patients with effective analgesia vs. time after intrathecal injection. àpatients delivering prior to loss of analgesia are included up to the time of delivery. Pain scores (VAS) Group CI-PCEA PCEA Time after induction of labour epidural analgesia (min) Figure 4 Pain scores over the first 8 h of study. 800 Table 4 Block characteristics and side-effects during study period. Values are n (proportion). haemodynamic and side-effect profiles after the CSE (Table 4). Mode of delivery, fetal weight and Apgar scores were also similar (Table 2). Discussion CI-PCEA group (n = 20) PCEA group (n = 20) p-value Pruritus 11 (55%) 9 (45%) Nausea 2 (10%) 1 (1%) Vomiting 2 (10%) 1 (5%) Shivering 9 (45%) 5 (25%) Hypotension 0 (0%) 2 (10%) Fetal bradycardia 0 (0%) 1 (5%) CI-PCEA, computer integrated-patient controlled epidural analgesia. PCEA, patient controlled epidural analgesia. In this study, we demonstrated that CI-PCEA, with a variable background infusion rate, did not increase the total amount of local anaesthetic used significantly. Its background infusion is only initiated when the parturient s labour pain increases and or as labour progresses. This potentially minimises local anaesthetic use in the initial stage of labour when pain scores are lower and epidural requirement is less [8]. The use of CI-PCEA was also associated with higher overall patient satisfaction scores. Several studies have demonstrated that background infusion improves pain scores [5, 11]. We did not detect any difference in the effectiveness of PCEA and CI-PCEA in terms of pain scores and breakthrough pain rate. Boselli et al. found that at higher background infusion rates of 6 ml.h )1 and 9 ml.h )1, there was a significant increase in overall local anaesthetic consumption without improvement in pain scores or patient satisfaction [3]. In spite of that, we found a higher satisfaction score for CI-PCEA. It is unclear whether the increased satisfaction was the result of the variable background infusion, although the trend towards less breakthrough pain may have contributed to this [10]. Journal compilation Ó 2006 The Association of Anaesthetists of Great Britain and Ireland 343

6 Y. Lim et al. Æ Computer integrated-patient controlled epidural analgesia Anaesthesia, 2006, 61, pages The lack of a background infusion in demand bolus PCEA has been associated with an increased incidence of breakthrough pain [5, 11]. The same could be inferred from the results of our study, which showed a 35% breakthrough pain rate in the demand only PCEA group. The CI-PCEA group demonstrated a 15% breakthrough pain rate. Unfortunately, our study was not powered to detect a significant difference, but a difference of 20% in the incidence of breakthrough pain may be clinically significant. In this initial study evaluating the use of CI-PCEA in parturients, three patients had breakthrough pain. Two patients had rapid progression of cervical dilation from 2 cm to 7 cm in the space of 2 and 3 h, respectively; with labour augmentation initiated after labour epidurals were given. Three boluses were activated within the first hour and background infusion was infusing at 15 ml.h )1. When patients demanded a subsequent epidural dose, the alarm alerted the anaesthetist to review the patients local anaesthetic requirement. The third patient had progression from cervical dilation of 3 cm to 9 cm over 2 h 45 min. In the last hour prior to breakthrough pain, the rapid progression of labour led her to demand four consecutive boluses, which activated the alarm and resulted in an anaesthetic intervention. The absence of basal infusion rate in our initial program may have resulted in recession of sensory block and successive demand boluses by the parturients may not have been sufficient following rapid progression of labour and sudden increase in labour pain. Therefore, the CI-PCEA algorithm should be refined and the role of CI-PCEA needs to be defined. This study showed that CI-PCEA can potentially combine the advantages of a demand only PCEA regimen (decreased local anaesthetic consumption) and a PCEA regimen with a background infusion (improved pain scores and decreased incidence of breakthrough pain). In conclusion, with CI-PCEA there is a similar local anaesthetic consumption compared to demand only PCEA, while increasing parturients satisfaction scores. Future research on this novel drug delivery program will include refining the clinical algorithm to provide seamless analgesia for parturients in labour and incorporating it into a PCEA pump which is responsive to the parturients needs. Acknowledgements The CI-PCEA was conceived and the programme written by Dr Alex T. Sia. CI-PCEA is currently in the process of being embedded into a commercial infusion pump by Dr Alex T. Sia. The study was supported by a grant from KK Research Centre, KK Women s and Children s Hospital, Singapore. The authors acknowledge the assistance of Dr Wendy Teoh in the recruitment of patients. References 1 Halpern SH, Muir H, Breen TW, et al. A multicenter randomized controlled trial comparing patient-controlled epidural with intravenous analgesia for pain relief in labor. Anesthesia and Analgesia 2004; 99: D Angelo R. New techniques for labor analgesia: PCEA and CSE. Clinical Obstetrics and Gynecology 2003; 46: Boselli E, Debon R, Cimino Y, et al. Background infusion is not beneficial during labor patient-controlled analgesia with 0.1% ropivacaine plus 0.5 microg ml sufentanil. Anesthesiology 2004; 100: Petry J, Vercauteren M, Van Mol I, et al. Epidural PCA with bupivacaine 0.125%, sufentanil 0.75 microgram and epinephrine for labor analgesia: is a background infusion beneficial? Acta Anaesthesiologica Belgica 2000; 51: Bremerich DH, Waibel HJ, Mierdl S, et al. Comparison of continuous background infusion plus demand dose and demand-only parturient-controlled epidural analgesia (PCEA) using ropivacaine combined with sufentanil for labor and delivery. International Journal of Obstetric Anesthesia 2005; 14: Missant C, Teunkenst A, Vandermeersch E, Van de Velde M. Patient-controlled epidural analgesia following combined spinal-epidural analgesia in labour: the effects of adding a continuous epidural infusion. Anaesthesia and Intensive Care 2005; 33: Panni MK, Segal S. Local anesthetic requirements are greater in dystocia than in normal labor. Anesthesiology 2003; 98: Capogna G, Celleno D, Lyons G, et al. Minimum local analgesic concentration of extradural bupivacaine increases with progression of labour. British Journal of Anaesthesia 1998; 80: Ruban P, Sia AT, Chong JL. The effect of adding fentanyl to ropivacaine 0.125% on patient-controlled epidural analgesia during labour. Anaesthesia and Intensive Care 2000; 28: Lim Y, Sia AT, Ocampo CE. Comparison of intrathecal levobupivacaine with and without fentanyl in combined spinal epidural for labor analgesia. Medical Science Monitor 2004; 10: Ferrante FM, Rosinia FA, Gordon C, Datta S. The role of continuous background infusions in patient-controlled epidural analgesia for labor and delivery. Anesthesia and Analgesia 1994; 79: Journal compilation Ó 2006 The Association of Anaesthetists of Great Britain and Ireland