Small dose of clonidine mixed with low-dose ropivacaine and fentanyl for epidural analgesia after total knee arthroplasty{dagger}

J. G. Förster* and P. H. Rosenberg

Department of Anaesthesiology and Intensive Care Medicine, Helsinki University Central Hospital, PB 340, FIN-00029 Helsinki, Finland

* Corresponding author. E-mail: johannes.forster{at}hus.fi

Accepted for publication July 4, 2004.


    Abstract
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. We studied whether a small dose of clonidine added to a ropivacaine–fentanyl mixture improves epidural analgesia without provoking side effects typically related to larger amounts of epidural clonidine.

Methods. In this randomized, double-blinded study, patients (≤85 yr, ASA I–III) underwent total knee arthroplasty (TKA) performed under spinal anaesthesia. After the operation, patients received an epidural infusion consisting of ropivacaine 2 mg ml–1 and fentanyl 5 µg ml–1 either without (Group RF, n=33) or with clonidine 2 µg ml–1 (Group RFC, n=36). The infusion rate was adjusted within the range 3–7 ml h–1.

Results. Average rate of infusion was slightly smaller in Group RFC than in Group RF (mean (SD) 4.7 (0.72) vs 5.2 (0.8) ml h–1, P=0.004). Compared with the RF group, patients in the RFC group required significantly less rescue pain medication, that is i.m. oxycodone (median (25th, 75th percentile) 0 (0, 7) vs 7 (0, 12) mg, P=0.027). Arterial pressure and heart rate were slightly lower in Group RFC throughout the study period (mean difference between the groups 5 mm Hg (P<0.002) and 3 min–1 (P=0.12), respectively). The groups did not differ statistically with respect to nausea, motor block, and sedation.

Conclusions. The small amount of clonidine added to the low-dose ropivacaine–fentanyl mixture reduced the need for opioid rescue pain medication after TKA. Clonidine slightly decreased arterial pressure and heart rate without jeopardizing haemodynamics. Otherwise, the side effect profiles were comparable in both groups.

Keywords: anaesthetic techniques, epidural, lumbar ; anaesthetics, local, ropivacaine ; analgesia, postoperative ; analgesics opioid, fentanyl ; clonidine


    Introduction
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Clonidine has been used as an adjuvant in regional anaesthesia in various settings,1 including postoperative epidural analgesia.23 Clonidine has been shown to improve analgesia when added to an epidural infusion of morphine after major abdominal surgery.4 Similarly, the co-administration of clonidine and local anaesthetic produced better analgesia than either drug alone.5 There are only a few studies in which clonidine has been added to an epidural infusion consisting of local anaesthetic and opioid.67 For example, thoracic epidural analgesia was improved by administering clonidine (20 µg h–1) together with a bupivacaine–fentanyl mixture after abdominal gynaecological surgery.7 The rationale to combine these drugs is that the component drugs may produce analgesia by additive or even synergistic mechanisms, and that the combination may allow reduced doses of each drug with correspondingly fewer dose-related side effects.8 In this study, we tried to take advantage of the analgesic potential of clonidine added to a low-dose epidural infusion consisting of ropivacaine (2 mg ml–1) and fentanyl (5 µg ml–1) after total knee arthroplasty (TKA). By using a low dose of clonidine (6–14 µg h–1), we tried to minimize the typical side effects of epidural clonidine such as hypotension, bradycardia, and sedation.19


    Methods
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
This prospective, randomized, controlled, and double-blinded study was approved by the local Institutional Ethics Committee and by the National Agency for Medicine. Written informed consent was acquired from all patients before the operation.

We enrolled 72 patients, aged ≤85 yr, ASA physical status I–III, undergoing primary, unilateral TKA. Exclusion criteria were contraindications to neuraxial anaesthesia and non-steroidal anti-inflammatory drugs, myocardial (NYHA class III–IV), renal or hepatic impairment, psychiatric illness, and BMI more than 36 kg m–2. Treatment allocation to the two study groups was by blocked randomization (closed envelope method). Block sizes were 20, 28, and 24. Randomization was performed by a third party not involved in the study. All investigators were kept unaware of the block sizes throughout the whole study period. All physicians and nurses involved in the treatment of the patients as well as the patients themselves were blinded concerning the study group allocation throughout the whole study period.

All patients received their normal morning medication before the operation, except angiotensin-converting enzyme inhibitors, diuretics, and diabetes mellitus medication. Pre-medication consisted of oral diazepam 5–15 mg. Intraoperative monitoring included pulse oximetry, ECG, non-invasive arterial pressure, and hourly urine output.

After local anaesthesia of the skin at the L2–L3 or L3–L4 inter-space, an epidural catheter (Portex Ltd, 18-gauge epidural minipack) was advanced 3–5 cm into the epidural space. A test dose of 3 ml of ropivacaine 7.5 mg ml–1 (Naropin®, AstraZeneca) was given through the epidural catheter. After that, spinal anaesthesia was given one lumbar inter-space below that of the epidural puncture site using 3–4 ml ropivacaine 7.5 mg ml–1. In case of early spinal block regression, anaesthesia was continued by intermittent injections of epidural ropivacaine 7.5 mg ml–1. As required, the patients received i.v. midazolam and fentanyl as additional pre-medication before the spinal and epidural punctures, or for sedation or pain during the operation (e.g. position related discomfort in the back or shoulder). Glycopyrrolate or atropine for bradycardia, as well as fluid challenge and ephedrine for hypotension were given as clinically indicated.

The patients received about 8 ml kg–1 of i.v. Ringer's acetate before the spinal anaesthesia. The intraoperative infusion therapy regimen consisted of Ringer's acetate and hydroxyethyl starch 60 mg ml–1. For thromboprophylaxis, we used s.c. dalteparin (Fragmin®, Pharmacia/Pfizer), 2500 IU 1 h after the administration of the spinal anaesthesia, followed by 2500 IU in the evening of the day of surgery, and after this 5000 IU day–1. A thigh tourniquet was used during surgery (320 mm Hg on average). Tranexamic acid (Caprilon®, Leiras, Finland) 0.5–1.0 g i.v. was given at the discretion of the surgeon for reduction of wound bleeding. After surgery, Ringer's acetate, hydroxyethyl starch, and packed red cells were given as clinically required.

Epidural infusion
After surgery, the patients were transferred to the recovery room. The epidural infusion was started when the sensory block had descended to Th10–L1 and motor function of the lower extremities started to recover (first voluntary contraction of thigh muscles). An anaesthesia nurse, who was not involved in the treatment of the patients otherwise, prepared the study solution. In the study group (RFC, n=36), the epidural mixture consisted of ropivacaine 2 mg ml–1, fentanyl 5 µg ml–1, and clonidine 2 µg ml–1 (Catapresan®, Boehringer Ingelheim, Ingelheim, Germany). A similar infusion without clonidine was used in the control group (RF, n=36). Initially, an epidural bolus of the study solution (5 ml) was given, after which the infusion was started at a rate of 5 ml h–1. The rate of the infusion was adjusted as required within the range 3–7 ml h–1. The infusion rate was increased in steps of 1–2 ml h–1 if the pain score was more than 3 evaluated on a visual analogue scale (VAS) from 0 (no pain) to 10 (worst pain imaginable). Similarly, it was increased on patients' request even though VAS was only 1–3. Infusion speed was decreased if motor block was grade 2–3, or in the presence of pronounced hypotension or bradycardia. Trigger values for hypotension and bradycardia were not standardized but were considered to be a matter of the individual clinical setting (taking into account the age of the patient, possible cardiovascular diseases, and the actual well-being of the patient). When the transition from spinal anaesthesia to epidural analgesia had been accomplished in a clinically satisfying way, the patients were transferred to the orthopaedic ward. The infusion was to continue until 12:00 on the first postoperative day (POD).

Study parameters
Study parameters were recorded in a double-blinded fashion by one of the authors (J.F.), a trained study assistant, or by the nurse on duty during night shift. Parameters were written down on the day of surgery at 18:00 and 24:00, and on the first POD at 06:00 and 12:00 (±30 min). Arterial pressure, heart rate, and pulse oximetry were recorded at these predetermined times. Patients were asked whether they had suffered from pruritus or from nausea and vomiting (PONV) at any time point between the previous and the present interview. PONV was scored on a three-step scale (no, feeling nauseated, retching or vomiting). At each of the four predefined interview times, the ‘worst’ or greatest value of pain intensity, motor block, and sedation observed after the previous interview was recorded. Pain intensity at rest and during motion (flexion of the knee, about 30°) was evaluated on the VAS for pain intensity (see previous section). Motor block was assessed using a modified Bromage scale10 (0=full flexion of knees and feet, 1=just able to move knees, 2=able to move feet only, 3=unable to move feet or knees). In addition, the degree of sedation was recorded (sleep=sound sleep at night, but no problem waking up, 0=awake, 1=snoozing and easy to wake up, 2=drowsy, 3=sleeping and difficult to wake up). At the end of the study infusion, each patient was asked whether he or she had felt drowsy to a disturbing extent at some time during the study period (i.e. subjective estimate by the patients: yes, no, cannot say). Finally, patients were asked about their overall satisfaction concerning the pain management regimen on a scale from 0 (worst) to 10 (best), and whether they would choose the same kind of pain treatment for a similar type of surgery in the future.

Supplemental analgesics and other drugs
All patients received propacetamol (ProDafalgan®, Bristol-Myers Squibb) 2 g i.v. at the time when the epidural infusion was started in the recovery room, and thereafter, paracetamol 1 g orally every 8 h. Rofecoxib (Vioxx®, MSD) 25 mg orally twice a day was started in the evening after the operation. Rescue pain medication was oxycodone 0.03 mg kg–1 i.v. (in the recovery room) or 0.05–0.07 mg kg–1 i.m. (after transfer to the orthopaedic ward). If two successive oxycodone doses, given shortly one after another, did not provide sufficient pain relief (VAS ≤3), the anaesthetist on call injected an epidural bolus of ropivacaine 7.5 mg ml–1 (3–7 ml) in order to test the function of the epidural catheter. PONV was treated as required with tropisetron (Navobane®, Novartis) 2 mg i.v., and additionally, as needed, with metoclopramide 10 mg i.v. Pruritus was treated with hydroxyzine (Atarax®, UCB) 10–25 mg orally.

Statistical analyses
Groups were compared with {chi}2-analysis or Fisher's exact test for categorical data, t-test, or two-way repeated-measures analysis of variance (RM ANOVA) for continuous, normally distributed data, and with Mann–Whitney U-test for non-parametric data. A P value <0.05 was considered statistically significant. We calculated the sample size based on estimations of the primary end point, that is the amount of the rescue medication i.m. oxycodone. Using previous experience (mean 11 mg, SD 8 mg) and assuming a mean reduction of 50% (5.5 mg) as clinically significant, 34 patients per group were considered necessary to detect statistical significance ({alpha}=0.05, two-sided, power=80%). All statistical calculations were performed using the StatView® for Windows® Version 5.0.1. computer program (SAS Institute Inc., Cary, NC, USA).


    Results
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Three patients, all from Group RF, had to be excluded completely from the statistical analysis. The data of another five patients (two in Group RF and three in Group RFC) were included in the statistical analysis until the time of their withdrawal. Reasons for withdrawal are presented in Table 1. The two groups were comparable concerning patient characteristics and intraoperative data (Table 2). Almost 80% of our patients received small amounts (Table 2) of i.v. fentanyl in the operating room, most of which was given as additional pre-medication before the spinal and epidural punctures. The groups did not differ concerning the pre-medication dose of oral diazepam, amounts of spinally administered ropivacaine, intraoperative bleeding, intraoperative tranexamic acid, and time from spinal anaesthesia to the start of the epidural infusion (data not shown).


View this table:
[in this window]
[in a new window]
 
Table 1 List of patients withdrawn and reasons for withdrawal.

 

View this table:
[in this window]
[in a new window]
 
Table 2 Patient characteristics, intraoperative data, and blood loss during the recovery room period.

 
Epidural infusion and analgesia
Three patients from Group RF presenting with slow regression of the primary anaesthesia were given reduced amounts of drugs at the start of the study infusion (twice epidural bolus 0 ml and once 3 ml, and initial infusion rate twice 4 ml h–1 and once 3 ml h–1, respectively). During the recovery room period, i.v. oxycodone was required by one patient in Group RFC (2 mg), and two patients in Group RF (one patient 2 mg, the other 3 mg). None of the patients needed an epidural bolus as rescue medication in the recovery room.

Compared with patients in Group RF, patients in Group RFC received smaller doses of epidurally infused ropivacaine and fentanyl (Table 3). At the same time, they required significantly less i.m. oxycodone as rescue medication (Fig. 1). The median amount (25th, 75th percentile) of i.m. oxycodone was 0 (0, 7) mg in RFC, compared with 7 (0, 12) mg in RF (P=0.027). One patient in Group RFC received an epidural bolus of ropivacaine as rescue medication compared with three patients in Group RF requiring altogether six boluses. Each epidural bolus improved pain relief, indicating an appropriate epidural position of the catheter. VAS pain scores at the interview time points did not differ statistically between the groups except at 24:00, in favour of the RFC group (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3 Data concerning the epidural infusion and analgesia.

 


View larger version (15K):
[in this window]
[in a new window]
 
Fig 1 Total amounts of i.m. oxycodone each patient received during the study period. Group RF without, Group RFC with clonidine. There is a statistically significant difference between the groups (MW-U, P=0.027), with smaller i.m. oxycodone requirements in Group RFC.

 
Haemodynamics and oxygen saturation
During the recovery room period, haemodynamic parameters did not differ between the groups and clonidine did not increase the incidence of hypotension (systolic arterial pressure <90 mm Hg), bradycardia (heart rate <50 beats min–1), or the use of ephedrine and atropine. At the four predetermined interviews, both mean arterial pressure and heart rate were lower in Group RFC than in Group RF (Fig. 2). The mean difference was 5 mm Hg for mean arterial pressure (P<0.002), and 3 min–1 for heart rate (P=0.12). When taking into account measurements apart from those made at the predetermined interviews, systolic arterial pressure less than 90 mm Hg and heart rate less than 50 beats min–1 were observed more often in the RFC group than in the RF group (Table 4). Pulse oximetry readings did not differ statistically between the groups. Some low oxygen saturation values (<92%) were encountered on the orthopaedic ward (Table 4).



View larger version (15K):
[in this window]
[in a new window]
 
Fig 2 (A) Mean arterial pressure (MAP) and (B) heart rate (HR) at the four interview times. Group RF without, Group RFC with clonidine. Times of predetermined interviews: 18:00 and 24:00 on the day of surgery, 06:00 and 12:00 on the first postoperative day. Symbols are mean (SD). With regard to MAP, there is a statistically significant difference between the groups (RM ANOVA, F=10.5, P<0.002).

 

View this table:
[in this window]
[in a new window]
 
Table 4 Occurrence of hypotension, bradycardia, desaturation, and side effects, as well as medication administered to treat side effects, on the orthopaedic ward.

 
Sedation, PONV, pruritus, and motor block
Table 4 presents data related to side effects and medication administered to treat side effects on the surgical ward. Degree of sedation assessed at the predefined interview times was similar in both groups. Sedation grade 3 was recorded once in both groups. Both cases occurred during the early morning. One third of all patients felt drowsy to a disturbing extent at some time during the study period (no inter-group difference). Eight patients in Group RFC suffered from vomiting (11 episodes), compared with one patient in Group RF. However, the differences in the occurrence of nausea and vomiting did not reach statistical significance. No obvious clinical relationship could be seen between the occurrence of hypotension and the occurrence of PONV. Less pronounced pruritus in Group RFC was recorded at 24:00. The degree of motor block at the predetermined interview times was comparable in both groups (median (25th, 75th percentile) 0 (1, 0) at 18:00 and 24:00, and 0 (0, 0) at 06:00 and 12:00). On the first postoperative morning, the physiotherapists could successfully perform their mobilization programme in 56 patients. In six patients (three in each group) the physiotherapists preferred to apply only a limited mobilization programme with the patients in bed because of a slight motor block. In two patients (one per group), physiotherapy was postponed because of pronounced motor block.

Technical problems and complications
Technical problems occurred in five patients. In one instance, the infusion pump had to be replaced (infusion interrupted for 40 min). In two cases, catheter occlusion was overcome by rearranging the attachment tapes of the epidural catheter. There was one disconnection and another case of occlusion, which led to withdrawal from the study (Table 1, no. 26 and no. 64).

One patient (76 yr, Group RFC) presented with postoperative confusion in the afternoon and evening of the 1st POD; that is after the study infusion was turned off. The next day, this patient was again cooperative but he had no recall of the previous day. One patient (70 yr, Group RFC) suffered from a bad headache during the morning of the 1st POD. The headache became worse and a computer tomogram of the head was taken in order to exclude major complications. Nothing exceptional was seen in the tomogram, and the headache disappeared after a small dose of i.v. diazepam.

Patient satisfaction
Overall satisfaction concerning the pain management regimen was good to excellent in both groups (Table 3). The majority of all patients said that they would choose the same kind of pain treatment for a similar type of surgery in the future (yes/no/cannot say; 33/1/1 in RFC vs 28/2/3 in RF).


    Discussion
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
In this study, clonidine augmented analgesia after TKA when added to a continuous low-dose epidural infusion of ropivacaine and fentanyl. Compared with the control group, patients in the clonidine group received, on average, smaller doses of epidurally infused drugs. At the same time, they required significantly less rescue medication. Enhanced analgesia was also reflected by the observation that in the clonidine group patients suffered less frequently from episodes of marked pain (VAS 4–6) and of breakthrough pain at rest (VAS 7–10) (Table 3). Nevertheless, pain relief was acceptable also in the control group; that is rescue pain medication was adequate, and patient satisfaction was good to excellent in both groups. Although statistically significant, the small difference in VAS scores at 24:00 in favour of the clonidine group is of limited clinical importance.

Clonidine produces dose-dependent spinal cord antinociception mainly through stimulation of {alpha}2-adrenoreceptors in the dorsal horn, mimicking the activation of descending inhibitory pathways.11112 When clonidine was used alone for continuous epidural analgesia, doses as high as 100–150 µg h–1 were required to obtain satisfactory analgesia.13 Such high doses cause hypotension, bradycardia, and sedation. In combination with either opioids or local anaesthetics, epidural clonidine has been used in single bolus doses of 75–800 µg or continuous infusion dosages of 20–50 µg h–1.2 However, these dosages of clonidine have also commonly been associated with hypotension, bradycardia, and sedation.2 Although usually well tolerated by patients, these side effects have been considered worrisome in the postoperative setting. Thoracic epidural analgesia was improved by administering clonidine at a rate of 18.75 µg h–1 together with a mixture of bupivacaine (5 mg h–1) and morphine (0.1 mg h–1) after hysterectomy,6 or by administering clonidine at a rate of 20 µg h–1 in a mixture of bupivacaine (6.25 mg h–1) and fentanyl (10 µg h–1) after abdominal gynaecological surgery.7 In these two studies,67 haemodynamic changes were observed. Compared with these above-mentioned studies, the clonidine dose used in our study (on average 9 µg h–1, i.e. approximately 220 µg days–1) was low. Unfortunately, the design of our study does not allow conclusions about whether clonidine supplemented analgesia just in an additive or, perhaps, a synergistic way.

Arterial pressure and heart rate were slightly lower in the clonidine group than in the control group. Similar modest haemodynamic changes have been described consistently in many previous studies utilizing clonidine for postoperative analgesia.167 Obviously, in addition to clonidine there are several other factors, which may have influenced arterial pressure and heart rate in our patients, such as fluid loss as a result of bleeding and diuresis, the amounts of administered fluids, and use of diuretics. Nevertheless, in the clinical sense, the lower arterial pressure and heart rate had no obvious deleterious impact on our patients.

The groups did not differ statistically concerning PONV and antiemetic drug consumption. This in spite of the fact, that patients in the clonidine group received fewer opioids. In fact, our results may suggest that clonidine may somehow be associated with the occurrence of nausea and vomiting (Table 4), which has been observed in animal research.14 In contrast, there are recent reports that {alpha}2-adrenergic agonists might be useful in the treatment of nausea and vomiting.1516 About one-third of our patients suffered from PONV, which is comparable with or even slightly less than in other reports using epidural analgesia with opioids (on average 42%17).

Sedation is a side effect frequently associated with the use of clonidine in postoperative analgesia, often in conjunction with opioids.12 Probably the low dose of clonidine used here does not contribute to sedation. The confusion that occurred in one of the elderly patients after the end of the study period may not be related to the study drugs, but rather to the type of surgery. Postoperative delirium in elderly patients is relatively common particularly after major orthopaedic surgery.18 The degree of motor block decreased over time. With the low-dose epidural infusions used here, motor block was slight overall, allowing early postoperative mobilization therapy in all but two patients.

Mixtures of ropivacaine and fentanyl, as well as of ropivacaine and clonidine have been shown to be stable for up to 30 days,19 and a combination of bupivacaine, morphine, and clonidine for up to 90 days.20 Triple combinations of local anaesthetic, opioid, and clonidine have been used as intrathecal boluses, either without2122 or with slight improvement of analgesia.23 Clonidine doses have ranged between 15 and 75 µg. In all these studies, the addition of clonidine caused significant hypotension. The advantage of such a single-shot technique may be that no spinal or epidural catheter is required but, on the other hand, hypotension seems to be a problem with the large intrathecal clonidine doses. The continuous administration of a ropivacaine–fentanyl–clonidine triple combination via a spinal catheter might be an interesting subject of future studies.

In conclusion, clonidine 2 µg ml–1 augmented analgesia when added to a continuous lumbar epidural infusion of ropivacaine 2 mg ml–1 and fentanyl 5 µg ml–1 after TKA at a rate of 3–7 ml h–1. However, the use of this low-dose epidural clonidine together with a multi-modal analgesic regimen consisting of paracetamol, rofecoxib, epidural ropivacaine, and fentanyl had no major clinical benefits. In particular, the reduced opioid consumption in the clonidine group did not translate into reduced side effects, for example PONV. The small clonidine dose did not cause sedation. The levels of arterial pressure and heart rate were slightly lower in the clonidine group as compared with the control group. Increasing the clonidine dose may further improve the quality of the pain management regimen, but it may also increase the risk of hypotension.


    Acknowledgments
 
This work was supported by a research fund granted by the Helsinki University Central Hospital. We are grateful to the colleagues and nursing staff of the Surgical Hospital, Helsinki University Central Hospital, for valuable assistance throughout the study, and to Pekka Tarkkila, MD, PhD, for critical advice during the preparation of the manuscript.


    Footnotes
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
{dagger} Presented in part at the Euroanaesthesia 2004 Meeting, Lisbon, Portugal, June 5–8, 2004. Back


    References
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Eisenach JC, De Kock M, Klimscha W. Alpha(2)-adrenergic agonists for regional anesthesia. A clinical review of clonidine (1984–1995). Anesthesiology 1996; 85: 655–74[ISI][Medline]

2 Armand S, Langlade A, Boutros A, et al. Meta-analysis of the efficacy of extradural clonidine to relieve postoperative pain: an impossible task. Br J Anaesth 1998; 81: 126–34[Abstract/Free Full Text]

3 Tryba M, Gehling M. Clonidine—a potent analgesic adjuvant. Curr Opin Anaesthesiol 2002; 15: 511–7[CrossRef]

4 Motsch J, Gräber E, Ludwig K. Addition of clonidine enhances postoperative analgesia from epidural morphine: a double-blind study. Anesthesiology 1990; 73: 1067–73[ISI][Medline]

5 Milligan KR, Convery PN, Weir P, Quinn P, Connolly D. The efficacy and safety of epidural infusions of levobupivacaine with and without clonidine for postoperative pain relief in patients undergoing total hip replacement. Anesth Analg 2000; 91: 393–7[Abstract/Free Full Text]

6 Mogensen T, Eliasen K, Ejlersen E, Vegger P, Nielsen IK, Kehlet H. Epidural clonidine enhances postoperative analgesia from a combined low-dose epidural bupivacaine and morphine regimen. Anesth Analg 1992; 75: 607–10[Abstract]

7 Paech MJ, Pavy TJ, Orlikowski CE, Lim W, Evans SF. Postoperative epidural infusion: a randomized, double-blind, dose-finding trial of clonidine in combination with bupivacaine and fentanyl. Anesth Analg 1997; 84: 1323–8[Abstract]

8 Solomon RE, Gebhart GF. Synergistic antinociceptive interactions among drugs administered to the spinal cord. Anesth Analg 1994; 78: 1164–72[ISI][Medline]

9 Bonnet F, Boico O, Rostaing S, et al. Postoperative analgesia with extradural clonidine. Br J Anaesth 1989; 63: 465–9[Abstract]

10 Bromage PR. A comparison of the hydrochloride and carbon dioxide salts of lidocaine and prilocaine in epidural analgesia. Acta Anaesthesiol Scand Suppl 1965; 16: 55–69[Medline]

11 Yaksh TL, Reddy SV. Studies in the primate on the analgesic effects associated with intrathecal actions of opiates, alpha-adrenergic agonists and baclofen. Anesthesiology 1981; 54: 451–67[ISI][Medline]

12 Murata K, Nakagawa I, Kumeta Y, Kitahata LM, Collins JG. Intrathecal clonidine suppresses noxiously evoked activity of spinal wide dynamic range neurons in cats. Anesth Analg 1989; 69: 185–91[Abstract]

13 De Kock M, Wiederkher P, Laghmiche A, Scholtes JL. Epidural clonidine used as the sole analgesic agent during and after abdominal surgery. A dose-response study. Anesthesiology 1997; 86: 285–92[ISI][Medline]

14 Japundzic-Zigon N, Samardzic R, Beleslin DB. Clonidine-induced emesis: a multitransmitter pathway concept. Pharmacol Res 1997; 35: 287–97[CrossRef][ISI][Medline]

15 Oddby-Muhrbeck E, Eksborg S, Bergendahl HT, Muhrbeck O, Lonnqvist PA. Effects of clonidine on postoperative nausea and vomiting in breast cancer surgery. Anesthesiology 2002; 96: 1109–14[CrossRef][ISI][Medline]

16 Khasawinah TA, Ramirez A, Berkenbosch JW, Tobias JD. Preliminary experience with dexmedetomidine in the treatment of cyclic vomiting syndrome. Am J Ther 2003; 10: 303–7[Medline]

17 Block BM, Liu SS, Rowlingson AJ, Cowan AR, Cowan JA jr, Wu CL. Efficacy of postoperative epidural analgesia: a meta-analysis. JAMA 2003; 290: 2455–63[Abstract/Free Full Text]

18 Williams-Russo P, Urquhart BL, Sharrock NE, Charlson ME. Post-operative delirium: predictors and prognosis in elderly orthopedic patients. J Am Geriatr Soc 1992; 40: 759–67[ISI][Medline]

19 Oster Svedberg K, McKenzie J, Larrivee-Elkins C. Compatibility of ropivacaine with morphine, sufentanil, fentanyl, or clonidine. J Clin Pharm Ther 2002; 27: 39–45[CrossRef][ISI][Medline]

20 Wulf H, Gleim M, Mignat C. The stability of mixtures of morphine hydrochloride, bupivacaine hydrochloride, and clonidine hydrochloride in portable pump reservoirs for the management of chronic pain syndromes. J Pain Symptom Manage 1994; 9: 308–11[ISI][Medline]

21 Grace D, Bunting H, Milligan KR, Fee JP. Postoperative analgesia after co-administration of clonidine and morphine by the intrathecal route in patients undergoing hip replacement. Anesth Analg 1995; 80: 86–91[Abstract]

22 Paech MJ, Banks SL, Gurrin LC, Yeo ST, Pavy TJ. A randomized, double-blinded trial of subarachnoid bupivacaine and fentanyl, with or without clonidine, for combined spinal/epidural analgesia during labor. Anesth Analg 2002; 95: 1396–401[Abstract/Free Full Text]

23 Sites BD, Beach M, Biggs R, et al. Intrathecal clonidine added to a bupivacaine-morphine spinal anesthetic improves postoperative analgesia for total knee arthroplasty. Anesth Analg 2003; 96: 1083–8[Abstract/Free Full Text]