Acta Pdiatrica ISSN 08035253

REGULAR ARTICLE

Comparison of a morphine and midazolam combination with morphine alone for paediatric displaced fractures: a randomized study
Chrystle Wille-Ledon1*, Hlne Chappuy (helene.chappuy@nck.aphp.fr)1,2,3*, Carole Giraud2, Jean-Marc Trluyer2,3,4, Grard Chron1
diatriques, Ho pital Necker Enfants Malades, Assistance Publique-Ho pitaux de Paris, Universite Paris Descartes, Paris, France 1.Service dUrgences pe Paris Descartes, Paris, France 2.EA3620, Universite de recherche Clinique et Centre dInvestigation Clinique Paris Centre, Paris, France 3.Unite pitaux de Paris, Universite Paris Descartes, Paris, France 4.Pharmacologie clinique, Groupe hospitalier Cochin-Saint-Vincent-de-Paul, Assistance Publique-Ho

Keywords Analgesia, Fractures, Orthopaedic injury, Pain management, Sedation Correspondence H Chappuy, Paediatric Emergency Department, Necker Enfants Malades Hospital, 149 rue de ` vres, 75743 Paris Cedex 15, France. Se Tel: +33-01-44-49-42-92 | Fax: +33-01-44-49-42-99 | Email: helene.chappuy@nck.aphp.fr Received 8 December 2010; revised 24 March 2011; accepted 7 April 2011. DOI:10.1111/j.1651-2227.2011.02311.x *Co-rst authors. Co-last authors. ClinicalTrials.gov Identier: NCT00416039.

ABSTRACT Aim: To compare the efcacy of sublingual midazolam with oral morphine versus
that of oral morphine with placebo in a paediatric population attending an emergency department (ED) with acute long-bone fractures. Methods: A sample of children aged 516 years with clinically deformed closed long-bone fractures was randomized to groups receiving either oral morphine (0.5 mg kg) sublingual placebo or oral morphine (0.5 mg kg) sublingual midazolam (0.2 mg kg). The main exclusion criteria were narcotic or benzodiazepine use, signicant head injury, multiple organ failure, femoral fracture and allergy. Pain scores were rated on a 100-mm visual analogue scale (VAS) at 0, 15, 30, 60, 90 and 120 min. Results: Fifty-eight children were enrolled (mean age: 10.5 years, SD 2.7). Fractures concerned the radius or ulna in 43 cases (74.1%), the humerus (22.4%) and the tibia or bula (3.5%). No signicant difference in VAS scores was observed between the two treatment arms (p = 0.72). Drowsiness was signicantly more frequent in the midazolam group (p = 0.007) during the rst 2 h after administration. No serious adverse event was observed. Conclusion: The analgesic performances of morphine and the combination of morphine with midazolam assessed by VAS were similar in children presenting at the ED with a long-bone fracture.

INTRODUCTION Long-bone fractures are a common reason for visits to paediatric emergency departments. These fractures are often very painful, and the management of this pain frequently requires the use of opiate analgesics administered intravenously or orally (13). Oral morphine treatment is effective for the management of pain caused by long-bone fractures in children, with rare minor side effects, such as mild drowsiness. Doctors often prefer this treatment to intravenous treatment. However, it generally takes more than an hour to achieve a satisfactory decrease [visual analogue scale (VAS) < 30] in pain (4). We aimed to decrease this delay by establishing a sedation analgesia protocol combining an opiate analgesic, morphine, with a sedative. We chose to combine oral morphine with midazolam, a short-acting benzodiazepine administered sublingually (5,6). Sublingual midazolam has been shown to be useful for the premedication during sedation before surgery, but no previous study has evaluated the efcacy and tolerance of sublingual midazolam used in combination with oral morphine for analgesia in displaced fractures in children. The aim of this study was thus to evaluate the combination of sublingual midazolam and oral morphine for the

management, in the emergency department, of pain owing to displaced fractures, focusing on the difference between VAS score at M0 (just before treatment administration) and that after 30 min (M30) and on safety of use.

MATERIALS AND METHODS Approval was obtained from the Necker-Enfant Malades ethical committee (Paris, France). We carried out a singlecentre, prospective, randomized, double-blind study comparing the combination of oral morphine and sublingual

Key notes
The objective of the study was to compare the efcacy of sublingual midazolam with oral morphine versus that of oral morphine with placebo in a paediatric population. No signicant difference in visual analogue scale scores was observed between the two treatment arms. Drowsiness was signicantly more frequent in the midazolam group (p = 0.007) during the rst 2 h after administration.

2011 The Author(s)/Acta Pdiatrica 2011 Foundation Acta Pdiatrica 2011 100, pp. e203e207

e203

Sedation analgesia in paediatric long-bone fractures

Wille-Ledon et al.

midazolam with the combination of oral morphine and a placebo. This study was carried out in the paediatric emergency department of Necker-Enfants Malades Hospital in Paris, France, under the supervision of Paris Descartes Clinical Research Unit. Children between the ages of ve and 16 years, accompanied by at least one of their parents and seen consecutively in the emergency department for a displaced fracture of a long bone requiring morphine analgesia, were included in the study once the parent or parents present had been informed about the procedure and had signed an informed consent form. The exclusion criteria were patients with ASA score >2, narcotic analgesia in the 6 h before arrival or benzodiazepine use in the 24 h before arrival, signicant head injury, multiple organ failure, femoral fracture, or allergy to opiates or benzodiazepine. Randomization and baseline data collection Patients and all study personnel except the investigative doctor were blinded to treatment assignment. Eligible patients were randomized 1:1 to receive midazolam or placebo. All patients were randomized using a computer-generated schedule with sequentially numbered containers prepared by the Clinical Research Unit. The treatment was administered by the investigative doctor, and the VAS was collected by a nurse blinded to the treatment assignment. Detailed information regarding therapy prior to the initiation of study drug, baseline demographics and illness were obtained at the time of enrolment. Drug administration The midazolam group (MDZ) received oral morphine (0.5 mg kg) and sublingual midazolam (0.2 mg kg). The control group (placebo) received the same dose of morphine orally and sublingual saline. Different syringes were used for morphine and midazolam or placebo. As midazolam has a very bitter taste, we added 0.5 mL of 30% glucose to make this drug more acceptable to the patients. The volume of liquid was the same for midazolam and placebo. The midazolam and placebo were prepared and administered by the doctor in the absence of the nurse responsible for pain evaluation, who was therefore blind to the contents of the syringe used. Thus, neither the patient nor the person responsible for noting VAS score in the observation notebook was aware of the group to which the patient had been assigned. Data collection On arrival at the emergency department (M0) and then 15 (M15), 30 (M30), 60 (M60), 90 (M90) and 120 (M120) min after treatment administration, pain was evaluated with a 100-mm VAS. We measured heart and respiratory rates, oxygen saturation, consciousness, nausea and vomiting. A nursing assistant, trained to recognize the secondary effects of sedation and who had successfully completed the paediatric resuscitation course (EPLS) of the European Resuscitation Council, remained with the child for the 2 h following administration of the treatment.

The primary endpoint was the difference between VAS score at M0 (just before treatment administration) and that at M30. The secondary endpoints were the differences in VAS score between M0, and M15, M60, M90 and M120. Patients with a VAS score of 30 or lower at the various times points were identied. We collected the following data to evaluate tolerance: respiratory rate (RR), oxygen saturation (SaO2), heart rate (HR), consciousness (drowsiness or agitation), nausea, vomiting and skin reactions. In all cases, the injured limb was provisionally immobilized with a pneumatic splint on the arrival in the emergency department. If reduction in the fracture was indicated in the emergency department, this procedure was carried out between 30 and 120 min after treatment administration, by an orthopaedic surgeon, in the presence of the paediatrician from the emergency department. A nitrous oxideoxygen mixture (Entonox) was administered to the child during the reduction procedure. Immediately after reduction, the child was asked whether he or she was willing to undergo the procedure again if necessary. Sample size calculation According to a previous study, the difference between VAS score at M0 (just before treatment administration) and that at M30 was )13 19 mm in the control group (4). We considered 15 mm to be a clinically signicant value for the difference between the two groups. Group sample sizes of 26 give a power of 81% to detect a difference of 15.0 between the null hypothesis that both group means are 13.0 and the alternative hypothesis that the mean of group 2 is 28.0 with known group standard deviations of 19.0 and with a signicance level (alpha) of 0.05 using a two-sided Students t-test, assuming that the true distribution is uniform. As a result of this estimation, it was decided to include 30 patients in each group. Statistical analysis The quantitative data are expressed as means standard deviation, and qualitative data are expressed as percentages with a 95% condence interval. Students t-tests were used to compare between the two groups, the means of the differences between VAS scores at different time points. Qualitative values were compared with Fishers exact tests. This analysis was carried out with NCSS 2007 software.

RESULTS Characteristics of the patients Sixty patients were randomized. However, owing to withdrawal of parental consent for two patients in group A, only 58 patients were analysed (see Fig. 1 for ow chart and Table 1 for patient characteristics). All the patients had an ASA physical status score of 1. The fractures were of the radius or ulna in 43 cases (74.1%), the humerus in 13 cases (22.4%) and the tibia or bula in two cases (3.5%). The fracture was displaced in 53 patients, 19 of whom underwent

e204

2011 The Author(s)/Acta Pdiatrica 2011 Foundation Acta Pdiatrica 2011 100, pp. e203e207

Wille-Ledon et al.

Sedation analgesia in paediatric long-bone fractures

80

60 randomized patients

70 60 50

Morphine Morphine+midazolam

VAS
Group A: Two parental consent withdrawn 28 patients received treatment Group B: 30 patients received treatment

40 30 20 10 0 0 20 40 60 80 100 120 140

Time (min)
Two hours follow up 28 attended Two hours follow up 30 attended

Figure 2 Progression of the mean value of the visual analogue scale score (SD) over the 120 min following treatment in the two groups.

Figure 1 Flowchart of the study population.

Table 2 Changes in VAS score Difference from initial VAS score (M0) Table 1 Description of the population MDZ group (n = 28) Age of the patients (years) (min; max) Weight (kg) (min; max) Sex ratio M F Number of displaced fractures VAS M0 (min; max) 10.5 2.73 5.215.6 37.2 13.3 1770 21 7 24 (86%) 69 18 30100 Placebo group (n = 30) 9.48 2.92 5.0814.2 33.4 12.1 1760 17 13 29 (97%) 63 25 15100 p value 0.18 0.43 MDZ (n = 28) M15 M30 M60 M90 M120 )15 )25 )35 )41 )44 15 28 26 28 28 Placebo (n = 30) )15 )22 )27 )41 )38 17 25 27 25 28 p value 95% CI of the mean 0.87 )26.53.5 0.77 )16.710.2 0.34)201.7.0 0.80)14.114.3 0.72)19.110.2 Number of patients with VAS score <30 MDZ (n = 28) 2 (7%) 5 (18%) 9 (32%) 11 (39%) 15 (54%) Placebo (n = 30) 5 (16%) 8 (27%) 10 (33%) 19 (63%) 14 (47%)

p value 0.42 0.53 0.93 0.11 0.79

MDZ = midazolam group; VAS = visual analogue scale. 0.17

MDZ = midazolam group; VAS = visual analogue scale.

that they would, and two children, one from each group, said that they would not. Safety The most frequently observed adverse effect was drowsiness, which was never more than moderate. It affected 24 children (40%; 95% CI: 2852%): 17 in the MDZ group (59%; 95% CI: 4177%) and seven in the placebo group (23%; 95% CI: 838%) (p = 0.007). All the somnolent patients were easy to wake with stimulation (level 3 of the Ramsay sedation score). No signicant relationship was evidenced between drowsiness and VAS score (data not shown). No serious adverse effect or hospitalization linked to treatment was observed. In the MDZ group, the adverse effects were vomiting (n = 1), a limited skin rash without oedema or signs of severity (n = 3) and moderate agitation for <30 min (n = 2). In the placebo group, the only adverse effect was vomiting (n = 2). No apnoea or desaturation was noted during the observation period in any of the children included.

reduction by an orthopaedic surgeon in the emergency department after M30: eight in the MDZ group and 11 in the placebo group. Periods of inclusion and follow-up were 21 months and 2 h, respectively. Efcacy There were 28 patients in the MDZ and 30 patients in the placebo group. The difference in VAS score between M0 and M30 did not differ signicantly between the two groups: )25 28 vs. )22 25 (p = 0.77). The difference remained non-signicant if we considered only children with an initial VAS score >30 or >70 (data not shown). There was no signicant difference between the two groups for mean VAS score at any of the times considered (see Fig. 2) or in the percentage of patients with a VAS score below 30 at any time point (see Table 2). After reduction in the fracture by an orthopaedic surgeon in the emergency department, 16 of the 19 children were asked whether they would agree to undergo the same procedure again. Fourteen children, seven from each group, said

DISCUSSION The oral administration of morphine combined with the sublingual administration of midazolam resulted in

2011 The Author(s)/Acta Pdiatrica 2011 Foundation Acta Pdiatrica 2011 100, pp. e203e207

e205

Sedation analgesia in paediatric long-bone fractures

Wille-Ledon et al.

analgesic effects similar to those obtained with morphine alone in terms of the time lag to action and efcacy. However, the combination treatment more frequently caused drowsiness. This drowsiness was mild, with verbal stimulation sufcient to wake the child. No other major cardiorespiratory effect or profound sedation was observed. Many published studies have reported an analgesic effect of benzodiazepines, either through their GABAergic activity or through interaction between sedation and analgesia (7). However, there are some conicting ndings concerning the consequences of GABA receptor activation according to receptor subtype and localization. In addition, very few data for children are available (810). Morphine derivatives and benzodiazepines are thus often used together. We choose to use midazolam as the benzodiazepine for this study, owing to its relatively short time to action (10 min following the sublingual administration of a dose of 0.3 mg kg, 2 min following IV administration of a dose of 0.1 mg kg), its short duration of action, the multiple possible routes of administration (oral, intranasal, sublingual, rectal, intramuscular and intravenous) and its safety of use provided that the conditions of use are respected (5,6). In this study, midazolam had a notable sedative effect: the number of drowsy children was signicantly higher in the MDZ than in placebo group. However, no effect on the pain measured on the VAS was observed. Our study provides no evidence of a relationship between analgesia and sedation. Morphine derivatives and benzodiazepines are frequently used together, but this is the rst study, to our knowledge, evaluating this biotherapy in the management of acute pain in children or adults and the interaction between sedation and analgesia. Other combinations of treatments have been evaluated for the management of pain associated with fractures in children. Friday et al. compared ibuprofen and paracetamolcodeine in patients aged ve to 17 years with acute traumatic pain in the extremities. The two treatments gave similar measurable analgesia in the rst hour of treatment, with minimal adverse effects (11). However, as this previous study used a colour analogue scale rather than the VAS, it is difcult to compare the results of the two studies. In another study comparing the efcacy for pain reduction of oxycodone and codeine in children aged four to 17 years with suspected forearm fractures, oxycodone tended to give greater pain reduction and less itching than codeine (12). Borland et al. (13) compared the efcacy of intranasal fentanyl and intravenous morphine for managing acute pain in children with deformed long-bone fractures. No difference between the two treatment groups was observed. The mean age of the patients, initial VAS scores and mean decrease in VAS score were similar to those of our population, but the decrease in VAS score seemed to be faster, with a nadir obtained after 20 min in this previous study. The use of oral glucose as an analgesic for newborns and infants is fully supported by the literature, but uncertainties include effectiveness beyond the newborn infant period (14). As midazolam and oral glucose were used together in

this study, our results suggest that oral glucose has little or no effect for children in situations such as bone fractures. Another source of potential bias in our study was the heterogeneity of the fractures as we included displaced and nondisplaced fractures. However, the number of displaced fractures was similar in the two groups. We chose to evaluate pain with VAS as this scale is valid for children aged 4 years and older. The discrimination of this score decreases with the age of the patients (15), but in our study, the ages of the patients in the two groups were similar. In conclusion, morphine and the combination of morphine and midazolam performed similarly in terms of analgesic efcacy in children with long-bone fractures. The combination of midazolam and morphine did not improve analgesia over that achieved with morphine alone in the population studied. Other combinations should be evaluated to decrease the percentage of children with insufcient analgesia on morphine alone following limb fractures. The two treatments that appear to be most effective in children for acute pain associated with long-bone fractures are ibuprofen and morphine. It would be informative to study the association of the two, because, as they have different mechanisms of action, their association may provide improved analgesia for children with broken legs or arms.

ACKNOWLEDGEMENTS This study was supported by a clinical research contract ` la Recherche Clinique) to Dr. Wille(Contrat dInitiation a ron from Assistance Publique Ho pitLedon and Prof. Che aux de Paris, France.

CONFLICT OF INTEREST There are no potential conicts of interest, real or perceived.

References
1. The National Agency for Health Accreditation and Evaluation. Evaluation and management strategies for acute pain in ambu diatr latory care of children aged 1 month to 15 years. Arch Pe 2001; 8: 42032. 2. Zempsky WT, Cravero JP; and the Committee on Pediatric Emergency Medicine and Section on Anesthesiology and Pain Medicine. Relief of pain and anxiety in pediatric patients in emergency medical system. Pediatrics 2004; 114: 134856. 3. Suc A, Wood C. Treatment of acute pain in children. In: Du JL, editor. Douleurs aigue casse s en situation durgence: des ` la de marche qualite . DeQuad Arnette 2004: 115 techniques a 22. ron G. Oral mor4. Wille C, Bocquet N, Cojocaru B, Leis A, Che diphine administration for childrens traumatic pain. Arch Pe atr 2005; 12: 24853. 5. Khalil S, Philbrook L, Rabb M, Wagner K, Jennings C, Chuang AZ, et al. Sublingual midazolam premedication in children: a dose responsive study. Paediatr Anaesth 1998; 8: 4615. 6. Geldner G, Hubmann M, Knoll R, Jacobi K. Comparison between three transmucosal routes of administration of midazolam in children. Paediatr Anaesth 1997; 7: 1039.

e206

2011 The Author(s)/Acta Pdiatrica 2011 Foundation Acta Pdiatrica 2011 100, pp. e203e207

Wille-Ledon et al.

Sedation analgesia in paediatric long-bone fractures

7. Jasmin L, Wu MV, Ohara PT. GABA puts a stop to pain. Curr Drug Targets CNS Neurol Disord 2004; 3: 487505. 8. Zeilhofer HU, Mo hler H, Di Lio A. GABAergic analgesia: new insights from mutant mice and subtype-selective agonists. Trends Pharmacol Sci 2009; 30: 397402. 9. Mirza NR, Munro G. The role of GABA(A) receptor subtypes as analgesic targets. Drug News Perspect 2010; 23: 35160. 10. Saari TI, Uusi-Oukari M, Ahonen J, Olkkola KT. Enhancement of GABAergic activity: neuropharmacological effects of benzodiazepines and therapeutic use in anesthesiology. Pharmacol Rev 2011; 63: 24367. 11. Friday JH, Kanegaye JT, McCaslin I, Zheng A, Harley JR. Ibuprofen provides analgesia equivalent to acetaminophencodeine in the treatment of acute pain in children with

12.

13.

14.

15.

extremity injuries: a randomized clinical trial. Acad Emerg Med 2009; 16: 7116. Charney RL, Yan Y, Schootman M, Kennedy RM, Luhmann JD. Oxycodone versus codeine for triage pain in children with suspected forearm fracture: a randomized controlled trial. Pediatr Emerg Care 2008; 24: 595600. Borland M, Jacobs I, King B, OBrien D. A randomized controlled trial comparing intranasal fentanyl to intravenous morphine for managing acute pain in children in the emergency department. Ann Emerg Med 2007; 49: 33540. Harrison D, Bueno M, Yamada J, Adams-Webber T, Stevens B. Analgesic effects ofvsweet-tasting solutions for infants: current state of equipoise. Pediatrics 2010; 126: 894902. von Baeyer CL. Childrens self-report of pain intensity: what we know, where we are headed. Pain Res Manag 2009; 14: 3945.

2011 The Author(s)/Acta Pdiatrica 2011 Foundation Acta Pdiatrica 2011 100, pp. e203e207

e207

Copyright of Acta Paediatrica is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.