Spinally and epidurally administered morphine is frequently associated with pruritus. Isolated case reports indicate that ondansetron may be effective in this context. This study aims to investigate the effectiveness of ondansetron to treat this side effect.


In a prospective, randomized, double-blind, placebo-controlled study, 100 patients with pruritus (> 4 on a visual analog scale, on which 0 represents no pruritus and 10 represents worst pruritus imaginable) after spinal or epidural administration of morphine, received either 8 mg ondansetron intravenously (ondansetron group) in 100 ml NaCl 0.9% or vehicle (placebo group). A decrease of more than 4 points on the visual analog scale 60 min after treatment was considered a success. Changes in levels of pain and sedation, hemodynamic values, and other side effects were checked regularly. The presence or absence of pruritus was assessed for the last time 24 h later.


The two groups were similar for demographic characteristics, the route of administration of morphine, and severity of pruritus at the beginning of the study. The ondansetron group showed a success rate of 70% versus 30% for the placebo group (P > 0.05). Among the successfully treated patients, three (9%) in the ondansetron group and six (40%) in the placebo group reported the recurrence of pruritus (P < 0.05). Among the successfully treated patients, none complained of residual pruritus 24 h later. No changes in pain or sedation levels were noted. Hemodynamic values remained stable, hemoglobin oxygen saturation did not decrease, and no other side effects were observed.


The administration of 8 mg ondansetron intravenously is an effective treatment for spinally or epidurally administered morphine-induced pruritus. In this clinical condition the treatment is safe and well tolerated.

THE efficacy of neuraxial morphine in the management of postoperative pain is well recognized. [1,2]Unfortunately, the use of morphine administered intrathecally or epidurally may be limited by a high incidence of side effects. The appearance of pruritus is frequent after this mode of application, especially after cesarean section. [3]This pruritus is often difficult to treat and responds poorly to conventional treatments except for naloxone and propofol, the two most effective means available to control this side effect. [4,5] 

Recently, a few case reports have shown the effectiveness of ondansetron in decreasing the intensity of pruritus linked with cholestatic liver disease [6,7]or associated with neuraxial morphine in adults [8]and in children. [9]We therefore undertook a prospective, randomized, doubleblind, placebo-controlled study to investigate the possible antipruritic properties of ondansetron in patients who received epidurally or intrathecally administered morphine.

We investigated 100 patients of American Society of Anesthesiologists physical status 1-3. These patients were of both sexes, were 18-70 yr old, weighed 40-100 kg, were scheduled for elective orthopedic surgery, and were presenting with pruritus induced by morphine administered epidurally (2 mg) or intrathecally (0.2 mg). All patients underwent regional anesthesia with plain bupivacaine (0.5%, given epidurally) or hyperbaric bupivacaine (0.5%, given spinally) without any other additives. None received intravenous opioids in the postoperative period, and none had received intra- or postoperative sedation with propofol or any other drug. Patients who had a pain score more than 40 on a visual analog scale (VAS; 0 = no pain and 100 = worst pain imaginable) received 2 g propacetamol given intravenously up to four times per day. Mefenamic acid (3 x 500 mg per day, given orally) was kept as a rescue treatment. Patients with an epidural catheter received bolus doses of 0.25% bupivacaine (5-8 ml) as necessary. Only pruritus appearing within the first 12 h after application of morphine was studied. Morphine was given epidurally at the beginning of surgery and spinally at the time of the puncture.

All patients gave informed consent before inclusion in the study. Approval of the institutional and ethical committee was obtained. Patients entered the study prospectively. Exclusion criteria were known allergy to ondansetron, history of any disease associated with pruritus, complaint of pruritus before surgery, increase in values on liver blood tests (alaninaminotransferase or aspartataminotransferase more than two times the normal range), and preoperative intake of histamine blockers. Three patients were excluded because of history of skin disease associated with pruritus, and one had an increase in values on liver blood tests. Medication was blinded and randomized by our pharmacy, which delivered coded vials of 96 ml saline, 0.9%, plus 8 mg ondansetron or 100 ml saline, 0.9%.

Patients complaining of pruritus were assessed with a VAS, on which 0 represented no pruritus and 10 represented worst pruritus imaginable. Patients who assessed the severity of the pruritus at a VAS score more than 4 were randomized and received in 10 min either 8 mg ondansetron in 100 ml saline, 0.9%, or vehicle (saline 0.9%) intravenously. After the end of the treatment, pruritus was evaluated by an independent research nurse every 10 min for 60 min.

A decrease of four points or more on the VAS, assessed 60 min after neuraxial administration of morphine, was considered a treatment success. A lack of such a decrease 60 min after the discontinuation of infusion of the study drug was considered a treatment failure. Patients whose treatment had failed received in an open fashion either 1 or 2 ml propofol, 1%, or 8 mg ondansetron intravenously at the discretion of the anesthesiologist. All successfully treated patients were checked one more time 24 h later. Other parameters documented in the course of the study were changes in the level of sedation using the Ramsay scale [10]; changes in the level of pain using a VAS (0 = no pain, 100 = worst pain imaginable); changes in blood pressure, heart rate, and oxygen saturation using a pulse oximeter; and appearance of headache or any other side effects. Mood changes were assessed by the presence of euphoria or irritability or no change during the study period.

A power analysis showed that 39 patients per group were necessary to obtain a [small beta, Greek] value of 80%, considering the success rate of the placebo group of 20% and anticipating a success rate of 50% in the ondansetron group. To increase the power, we included 50 patients per group. Results are expressed as mean +/− SD. Demographic and morphometric characteristics, surgical data, and initial pruritus score were compared using the Mann-Whitney U test. The influence of the initial pruritus score on the success rate was assessed using the Mann-Whitney U test, the success rate between the groups was assessed by chi-square analysis, and relapses within 24 h and influence of sex on treatment success was assessed using Fisher's exact test. For all determinations, a probability value < 0.05 was considered significant.

Five hundred twenty-six patients were enrolled in the study, which gives an incidence of epidurally or spinally administered morphine-induced pruritus of 19%. Both groups were comparable regarding demographic characteristics, surgical data, and route of administration of opioids (epidural vs. intrathecal;Table 1). Three patients in the ondansetron group withdrew during the study evaluation because they wanted to be sure not to receive the placebo and subsequently refused to score the pruritus intensity. Three other patients from this group received metoclopramide (one received 10 mg and two received 10 mg twice, both intravenously) for emesis. In the placebo group, three patients received metoclopramide 10 mg twice, given intravenously for emesis.

Table 1. Characteristics of Patients and Surgical Data

Table 1. Characteristics of Patients and Surgical Data
Table 1. Characteristics of Patients and Surgical Data

The pruritus appeared 3.5-7.0 h after neuraxial administration of morphine. In more than 90% of patients it involved the thorax and face.

The treatment success rate was significantly greater in the ondansetron group (33 of 47 patients, 70%) than in the placebo group (15 of 50 patients, 30%)(P < 0.05). The success rate was not influenced by the sex in either group (Table 2). The intensity of pruritus as measured on the VAS at the beginning of the study was similar in both groups: 5.8 +/− 0.4 (range, 4.4-10.0) versus 5.4 +/− 0.3 (range, 4.3-8.5) in the ondansetron and placebo groups, respectively (Table 2). Only one patient in the ondansetron group had an initial pruritus score of 10; he was treated successfully. We did not find any statistically significant correlation between the initial value of pruritus score and the treatment success. Among the successfully treated patients, none complained of residual pruritus 24 h after the completion of the study period. Among those, 3 of 33 (9%) in the ondansetron group and 6 of 15 (40%) in the placebo group (P < 0.05) reported the recurrence of slight pruritus during the subsequent night. The pruritus was treated successfully with 8 mg ondansetron given intravenously to five patients (one in the ondansetron group and four in the placebo group) and with 10 mg propofol given intravenously to four patients (two in the ondansetron group and two in the placebo group). In the ondansetron group, recurrence of pruritus appeared 5-7 h after the drug was given.

Table 2. Results

Table 2. Results
Table 2. Results

Among the unsuccessfully treated patients in the ondansetron group, six were treated successfully with 10 or 20 mg propofol given intravenously, two needed a continuous infusion of 1 mg [middle dot] kg-1[middle dot] h-1propofol, two did not respond at all, and four did not wish to undergo any other treatment. In the placebo group, 17 patients were treated successfully with 8 mg ondansetron given intravenously and 5 with 10-20 mg propofol given intravenously, 2 did not respond at all, and 11 did not wish to undergo any other treatment.

The sedation scale and the pain score did not change during the study period in either group. Hemodynamic changes were similar in both groups and were nonsignificant. Neither respiratory rate nor hemoglobin oxygen saturation measured by pulse oxymetry changed in either group. No headaches or mood changes were noted.

The incidence of pruritus after intrathecally or epidurally administered morphine is high and may limit its application. [3]This study shows that 8 mg ondansetron given intravenously increases the success rate of treatment significantly compared with placebo. This success was achieved without any change in the levels of postoperative pain and sedation, hemodynamic modifications, or the presence of other side effects.

The dose of 8 mg of ondansetron was chosen because an open preliminary pilot study (unpublished data) showed that this dose was effective and well tolerated. The study period was limited to 60 min because we noted in our preliminary pilot study that positive treatment effects nearly always occurred within the first 30 min. This is in accordance with the few available data dealing with this question. Crighton et al. [11]successfully treated two obstetric patients who had spinally administered morphine-induced pruritus with 4 mg ondansetron; in these two patients, pruritus was improved within 30 min. Larijani et al. [8]treated two patients unresponsive to diphenhydramine, one with 4 mg and one with 8 mg ondansetron given intravenously. The pruritus in these three patients improved greatly within 10 min. Another patient did not respond to 12 mg ondansetron given intravenously. Two pediatric patients had positive responses after 2 and 4 mg ondansetron given intravenously. [9]Some patients with cholestatic-associated pruritus also were successfully treated with 8 mg ondansetron given intravenously. [7,12] 

The high treatment success rate observed in the current study (70%) is in agreement with the published data (83%)[8,11]and is comparable to the success rates observed with propofol or naloxone. [5,13]The recurrence rate within the first 24 h (9%) also is comparable to that reported in the literature (13%). [8,9,11]Opioid-induced pruritus may be extremely difficult to manage, and approximately 10-15% of patients remain unresponsive to propofol or naloxone. [5,13]In the current study, 15% also were resistant to the administration of ondansetron and subsequently propofol. The incidence of morphine-induced pruritus is very high and particularly difficult to treat in obstetrics, [3]but in the current study gender did not influence the success rate in either group. Side effects associated with ondansetron, such as headaches, abdominal pain, [14]and cardiac dysrhythmia, [15]were not observed.

The success rate observed in the placebo group (30%) is higher than that noted after infusion of Intralipid (Pharmacia & Upjohn, Dubendorf, Switzerland) in the same clinical context [5]but remains in the range of that quoted in the literature for postoperative nausea and vomiting. [16,17]The large variation of the placebo effect in the context of pruritus may be explained by the importance of subjective feelings and psychologic factors.

The absence of any change in level of pain observed in this trial is consistent with the results of Petersen-Felix et al., [18]who showed that, in healthy volunteers, experimental pain induced with heat, cold, mechanical pressure, and electrical stimulation were not altered by 8 mg ondansetron given intravenously. The authors also demonstrated that the analgesia effect produced by alfentanil was not antagonized by ondansetron. [18] 

Intrathecally and spinally administered opioids have been shown to be associated with pruritus and the related scratching behavior. [1,19]These symptoms typically spread rostrally from the site of administration, followed by itching on the trunk that reaches the face approximately 2-4 h later, which provokes itching on the nose and particularly around the eyes. [19]The precise cause of this pruritus is not completely clear. The most convincing hypothesis remains that this pruritus is caused by a direct, excitatory effect of morphine on the dorsal horn. [19]This concept is supported by the facilitatory action of morphine on the nonnociceptive neurons in the posterior horn [20]and indirectly by the beneficial effects of propofol, which is known to possess a strong depressant effect on the dorsal horn. [21]The beneficial effects of ondansetron raise a new question: Does the 5-hydroxytryptamine 3 (5-HT3) receptor play a role in this pruritus? Hamon et al. [22]found a dense concentration of 5-HT3receptors in the dorsal part of the rat spinak cord, particularly in the superficial layers of the dorsal horn. These results are in accordance with those of Waeber et al., [23]who evidenced a relatively high density of 5-HT3receptors in the nucleus of the spinal tract of the trigeminal nerve in the mouse medulla. It is noteworthy that this nucleus is for the face what the superficial layers of the dorsal horn are for the body. [22,23]Thomas et al. [24]emphasized in this context the role of the medullary dorsal horn, the brain stem homologue of the dorsal spinal cord, demonstrating that microinjections of morphine in this anatomic region produce a dose-dependent, naloxone-reversible pruritus on the face of the monkey. Moreover, propofol, which has a strong depressant effect on the spinal dorsal horn, [21]is, among different anesthetic agents, the one that possesses the most potent noncompetitive inhibition of 5-HT3-inducedinflux of14C-guanidinium, pointing toward a specific interaction with the 5-HT3receptor channel. [25]These observations suggest that the 5-HT3receptor is, to a certain degree, implicated in the development of the pruritus associated with the application of neuraxial opioids.

This study shows that 8 mg ondansetron administered intravenously is effective to treat spinally or epidurally administered morphine-induced pruritus. Ondansetron is well tolerated, does not change the levels of analgesia, and is not associated with side effects. More studies are needed to investigate the dose-response relation of ondansetron in this context and to evaluate the potential synergism of ondansetron added to other drugs, such as propofol or naloxone, in the treatment of refractory cases of morphine-induced pruritus.

Cousins MJ, Mather LE: Intrathecal and epidural administration of opiods. Anesthesiology 1984; 61:276-310
Jacobson L: Intrathecal and extradural narcotics, Advances in Pain Research Therapy. Edited by Benedetti C. New York, Raven Press, 1984, pp 199-236
Harrison DM, Sinatra R, Morgese L, Chung JH: Epidural narcotic and patient-controlled analgesia for post-cesarean section pain relief. Anesthesiology 1988; 68:454-7
Vedrenne JB, Esteve M, Guillaume A: Prevention par la naloxone des effets secondaires de l'analgesie peridurale par morphine pour douleur cancereuse. Ann Fr Anesth Reanim 1991; 10:98-103
Borgeat A, Wilder-Smith OHG, Saiah M, Rifat K: Subhyponotic doses of propofol relieve pruritus induced by epidural and intrathecal morphine. Anesthesiology 1992; 76:510-2
Schworer H, Ramadori G: Treatment of pruritus: A new indication for serotonin type 3 receptor antagonists. Clin Invest 1993; 71:659-62
Schworer H, Hartmann H, Ramadori G: Relief of cholestatic pruritus by a novel class of drugs: 5-Hydroxytryptamine type 3 (5-HT3) receptor antagonists: Effectiveness of ondanserton. Pain 1995; 61:33-7
Larijani GE, Goldberg ME, Rogers KH: Treatment of opioid-induced pruritus with ondansetron: Report of four patients. Pharmacotherapy 1996; 16:958-60
Arai L, Stayer S, Schwartz R, Dorsey A: The use of ondansetron to treat pruritus associated with intrathecal morphine in two paediatric patients. Paediatr Anaesth 1996; 6:337-9
Ramsay MAE, Savage TM, Simpson BRJ, Goodwin R: Controlled sedation with alphaxalone-alphadolone. BMJ 1974; 2:656-9
Crighton IM, Hobbs GJ, Reid MF: Ondansetron for the treatment of pruritis after spinal opioids. Anaesthesia 1996; 51:199-200
Schworer H, Ramadori G: Improvement of cholestatic pruritus by ondansetron. Lancet 1993; 341:1277
Saiah M, Borgeat A, Wilder-Smith OHG, Rifat K, Suter PM: Epidural-morphine-induced pruritus: Propofol versus naloxone. Anesth Analg 1994; 78:1110-3
Baber N, Palmer JL, Frazer NM, Pritchard JF: Clinical pharmacology of ondansetron in postoperative nausea and vomiting. Eur J Anaesth 1992; 9:11-8
Baguley WA, Hay WT, Mackie KP, Cheney FW, Cullen BF: Cardiac dysrhythmias associated with the intravenous administration of ondansetron and metoclopramide. Anesth Analg 1997; 84:1380-1
Borgeat A, Wilder-Smith OHG, Saiah M, Rifat K: Subhypnotic doses of propofol posses direct antiemetic properties. Anesth Analg 1992; 74:539-41
Beecher HK: the powerful placebo. JAMA 1995; 159:1602-6
Petersen-Felix S, Arendt-Nielsen L, Bak P, Bjerring P, Breivik H, Svensson P, Zbinden AM: Ondansetron does not inhibit the analgesic effect of alfentanil. Br J Anaesth 1994; 73:326-30
Ballantyne JC, Loach AB, Carr DB: Itching after epidural and spinal opiates. Pain 1988; 33:149-60
Belcher G, Ryall RW: Differential excitatory and inhibitory effects of opiates on non-nociceptive and nociceptive neurons in the spinal cord of the cat. Brain Res 1978; 145:303-14
Cavazzuti M, Porro CA, Barbieri A, Galetti A: Brain and spinal cord metabolic activity during propofol anaesthesia. Br J Anaesth 1991; 66:490-5
Hamon M, Gallissot MC, Menard F, Gozlan H, Bourgoin S, Verge D: 5-HT3 receptor binding sites are on capsaicin-sensitive fibres in the rat spinal cord. Eur J Pharmacol 1989; 164:315-22
Waeber C, Dixon K, Hoyer D, Palacios JM: Localisation by autoradiography of neuronal 5-HT3 receptors in the mouse CNS. Eur J Pharmacol 1988; 151:351-2
Thomas DA, Williams GM, Iwata K, Kenshalo DR, Dubner R: The medullary dorsal horn: A site of action of morphine in producing facial scratching in monkeys. Anesthesiology 1993; 79:548-54
Barann M, Gothert M, Fink K, Bonisch H: Inhibition by anaesthetics of14C-guanidinium flux through the voltage-gated sodium channel and the cation channel of the 5-HT3 receptor of N1E-115 neuroblastoma cells. Nauryn Schmiedebergs Arch Pharmacol 1993; 347:125-32