The safe and effective treatment of pain following daycase paediatric surgery, remains a challenge [1-3]. The misconception that children feel less pain than adults, has been proved wrong . However fear of side effects, such as respiratory depression, prevents the use of powerful opioids under ambulatory conditions . For the child, acceptability of the route of administration is as important as the analgesia provided and pleasant tasting oral formulations are preferred to injections or suppositories [1,5].
Dental surgery is a common paediatric day-case procedure and is one of the standard models used for studying the efficacy of minor analgesic agents [3,6]. The severity of the post-operative pain is related to the number of teeth being extracted. By studying children having six or more extractions , we felt that an effective clinical model was available.
Tramadol hydrochloride (Tramal®, Grünenthal, Germany) was developed as a centrally acting analgesic with spinal and supraspinal sites of action [7,8]. In young adults, a single oral dose has a bioavailability of 70% with the peak serum concentration being reached at 2 h. Clinical analgesia becomes apparent from 20 min and is maintained for 6-9 h. It has not yet been used in the paediatric post-extraction pain model.
Tramadol has an affinity for opioid receptors that is similar to that of codeine and the adverse events are typically central in nature, such as dizziness, nausea and vomiting . The incidence of these is low, vomiting being reported in less than 1% of adult patients . A study in children aged 2-12 years using intramuscular (i.m.) tramadol, found no significant respiratory effects . This suggests that this agent would be effective when given as a pre-operative analgesic to ensure pain relief on awakening from surgery.
Midazolam (Dormicum, Roche, Switzerland) is a short acting water soluble benzodiazepine. It is widely used as an enteral route pre-anaesthetic anxiolytic for children [11,12]. Benzodiazepines are known to potentiate the actions and adverse events of opioids . The effects of combining midazolam and tramadol in the premedication of children is unknown.
We therefore took children aged 4-7 years, having six or more teeth extracted and investigated the effect of oral tramadol 1.5 mg kg−1 plus midazolam 0.5 mg kg−1 compared with midazolam 0.5 mg kg−1 alone on pre-anaesthetic behaviour, respiratory and cardiovascular status, recovery rate and post-operative analgesia.
Sixty ASA class I children (32 male, 28 female) were included in a double-blind, randomized, placebo-controlled trial after written informed consent of the parents and approval by the University Ethics committee had been obtained. Children aged 4-7 years old, had to undergo six or more dental extractions (maxillary and mandibular) under general anaesthesia.
Children in group A were premedicated with oral midazolam 0.5 mg kg−1 (maximum dose 7.5 mg) together with tramadol drops 1.5 mg kg−1. Group B children received midazolam 0.5 mg kg−1 (maximum dose 7.5 mg) and placebo (normal saline in drop form). Both premedication drugs were administered orally 30 min before induction of anaesthesia. Neither, the anaesthetist, who did the anaesthetic assessments, nor the research sister who did the post-anaesthetic recovery and pain assessments, were informed which study medication was administered.
Acceptance of oral administration was classified according to whether it was good, with no defensive reactions; moderate, with defensive reactions and weeping; and poor, with refusal and weeping. Behaviour was scored by the research sister according to whether the child was very distressed, weeping; distressed, not weeping; calm, indifferent, not weeping; and asleep. This was carried out before administration of the study drugs, 20 min after, and immediately before induction of anaesthesia (30 min). Level of sedation was assessed at the same time intervals according to a modified Ramsay sedation scale . Children were scored whether they were fully awake, orientated; drowsy; eyes closed but rousable to verbal command; and eyes closed, unrousable to mild physical stimulation.
Heart and respiratory rates, blood pressure (systolic, diastolic and mean) and oxygen saturation were monitored prior to administration of drugs, 20 min thereafter and immediately before induction of anaesthesia using a dinamap automatic blood pressure apparatus (model 1846 SX, Critikon, USA).
General anaesthesia was induced with a standard sevoflurane inhalation technique, using a step-wise increase in concentration up to 8%. Acceptance of mask induction was recorded by the anaesthesiologist by evaluating body movement, crying and ease of induction. After nasotracheal intubation, children were allowed to breathe spontaneously. Anaesthesia was maintained on 1.5-2.0 MAC sevoflurane in nitrous oxide (60%) and oxygen (40%). When necessary, breathing was assisted to maintain an end-tidal carbon dioxide concentration of between 4 and 5%. Monitoring in theatre consisted of standard temperature, blood pressure, electrocardiogram, pulse oximetry and capnography.
On completion of surgery, children were transferred to the recovery room where they were continuously monitored by a research sister trained in recovery, pain and behaviour assessment. A 35% oxygen mask was placed on the child's face until they awoke sufficiently to reject it. Post-operative recovery was assessed according to the Aldrete post-anaesthetic recovery score . All the parameters were assessed on admission, after 15 min, 30, 60 and 120 min. The score ranges from zero for an unresponsive, immobile child requiring airway maintenance, to 10 for a fully recovered child.
Post-operative pain in this study was assessed using the Oucher faces pain scale  (OFPS, score 1-6) and a behavioural-cardiovascular checklist (Hannallah  objective pain scale - HOPS, score 0-14). The Oucher faces pain scale comprises six faces depicting increasing gradations of pain severity, from 'no pain' on the bottom face to 'most pain' on the top face. The chosen faces were converted to a numerical visual analogue score (VAS) where 0 represents 'no pain' and 100 represents 'most pain possible.' The research sister, mother and child were asked to assess the pain at 15, 30, 60, 90 and 120 min post-operatively.
Oxygen saturation (Ohmeda Biox III, Bioximetry technology Inc, Boulder, CO, USA) was assessed on admission and at the same above time points. All assessments were continued for 2 h in the recovery room.
Post-operative analgesia was administered as paracetamol syrup 120 mg (5 mL), as needed. This was given when the OFPS was 3 or above or the HOPS score was 4 or above. Records of adverse experiences were made.
Results are presented as mean values±standard deviations. The χ2-test was used for the behavioural parameters, a P-value < 0.05 was taken as significant.
For the blood oxygen saturation a drop below 90% would be clinically relevant, hence an n=26 per treatment group was regarded as sufficient to detect a difference. A continuous variable in this study was the blood oxygen saturation. The difference between the treatment groups with regard to the mean minimum value of the blood oxygen saturation was tested exploratorily using the two-sample t-test (two-tailed). As this was not the primary variable of the study, the resulting P-value is to be interpreted in the exploratory sense. For the efficacy end point (pain) the hypothesis of equal rates was analysed by Fisher's exact test at the level α=0.05; 95% confidence intervals were calculated for the respective frequencies per treatment and for their difference.
Children in the two groups were similar with respect to age, height, weight, gender distribution and dental extractions (Table 1). The blood pressures, heart and respiratory rates showed no differences at any time before or after anaesthesia. The mean haemoglobin oxygen saturation levels immediately before induction were 98.2%, SD 1.3 (group A) and 97.7%, SD 1.2 (group B) (P>0.05). The mean haemoglobin oxygen saturation levels immediately before induction of anaesthesia were 98.2%, SD 1.3 (group A) and 97.7%, SD 1:2 (group B) (P>0.05). At this time one child (3.2%) in group A had an oxygen saturation of 94%, none had levels between 94 and 97%; in group B five children (17.2%) had oxygen saturation levels between 94 and 97%. Intra-operative oxygen saturations were all above 97%. During the 2 h in the recovery room, eight children (25.8%) in group A had oxygen saturation levels between 94 and 97%, compared with nine children (31.0%) in group B (P>0.05). These values were in actively moving children. All of the other post-anaesthetic levels were more than 97%.
In both groups, oral administration of the drugs was well accepted with no differences between the two groups.
Regarding pre-operative sedation and behaviour as reflected by specific scores at different time intervals, no significant differences were found between the two groups. Immediately before mask induction of anaesthesia, 93.5% of group A children were drowsy, while 3.2% of children were asleep but rousable to verbal command. In group B, 93.1% of children were drowsy, while 3.4% were asleep but rousable to verbal command. During induction of anaesthesia, 9.7% of group A children demonstrated mild weeping compared with 6.9% of group B children. Ease of mask induction was recorded as very good in 93.5% of group A children compared with 86.2% in group B (P>0.05).
Post-operative recovery (Aldrete score 10)  was reached after 48.8 min, SD 32.6 in group A children, compared with 36.4 min, SD 29.6 in group B children (P>0.05). However, the oxygen masks were tolerated for 10 min less than these times in both groups. The development of the post-operative pain, reflected by the Oucher scale, is shown in the Figs. 1 and 2. The maternal and research sister's assessments of pain were very similar to those of the children. Therefore only the children's self assessments are shown in Fig. 1. It must be taken into account that these are true values, i.e. derived from data recorded prior to the administration of post-operative paracetamol as well as from results obtained thereafter. Therefore, a second evaluation was conducted on the basis of the data recorded prior to the administration of rescue medication. In each child receiving paracetamol, all assessments recorded thereafter, were replaced by the last observation carried forward (LOCF). This approach is seen in Fig. 2. The tramadol group of children had significantly less pain than those in the placebo group, at all time intervals (P<0.05). At 60 min, the tramadol group's pain score is half that of the placebo group's. This decreases to one third at 90 and 120 min
In group A, 19.4% (six) of the children received postoperative rescue paracetamol compared with 82.8% (24) in group B (P<0.05). The efficacy of the rescue medication can be seen from the Hannallah objective pain scale 10 scores, Table 2. This shows the results at each scheduled determination time. No differences could be demonstrated (P>0.05).
In order to exclude the effect of post-operative analgesia (LOCF), the maximal mean results up to the administration of paracetamol were also compared. This showed a mean score of 6.0, SD 3.0 in the tramadol group, compared with 7.5, SD 3.0 in the placebo group (P=0.05).
With regard to the entire study documentation, the occurrence of an adverse experience (mild post-operative vomiting) was recorded in one group B patient only. The event corresponded to a single episode with a duration of 2 min.
The emotional distress generated by an operating theatre experience is aggravated by the severity of the pain that develops on recovering from anaesthesia [1, 17]. In children this distress is associated with behavioural disorders on returning home and a negative attitude to future medical procedures [18, 19]. These may vary from mild to severe. The incidence is decreased by day-stay surgery, suitable sedative-anxiolytic premedication and adequate analgesia. Minor surgery, such as dental extractions, carried out as day cases seldom warrant the use of powerful opioids with their attendant adverse effects [4, 10]. Numerous minor analgesics have been used for post-dental extraction pain, preferably via the oral route [20, 21]. However, due to the possibility of deleterious interactions any sedative-analgesic combination requires clinical testing prior to general use.
With regard to pre-anaesthetic sedation, both groups were equally drowsy and peaceful, resulting in a smooth mask induction of anaesthesia in the majority of the children. Tramadol had no obvious effect on the pre-anaesthetic behaviour of the children, allowing the anxiolytic property of midazolam to be successfully used [11, 12]. The pharmacodynamic actions of tramadol did not result in deleterious respiratory or cardiovascular effects. No differences were seen between the tramadol and placebo groups of children. All the vital parameters were similar.
Despite the potentially additive effects of a benzodiazepine plus an opioid receptor acting agent on respiratory depression , no such depression could be demonstrated. Pre-anaesthetic and post-anaesthetic oxygen saturation levels were the same for both groups at mean values of 97-98%. This is in agreement with previous work on tramadol in children [9, 10].
During anaesthesia, no adverse events were seen and the post-operative incidence of adverse events was minimal, only one placebo group child vomited. Hence the addition of tramadol was shown to have no safety implications in the 31 children, aged 4-7 years, studied. Full active recovery took 12 min longer in the tramadol group (48.8 min vs. 36.4 min) than in the placebo group. This was not statistically significant and during this extra time the children did not require airway maintenance. This non-significant time delay to active awakening has been noted in studies utilizing other effective analgesic techniques . Pain prevention decreases the central nervous system activation resulting from nociceptive stimuli.
Single dental extractions in children are not associated with significant pain [20,21]. However, the pain increases as the number of extractions increases, more than five teeth being associated with significant discomfort [3,6]. Our children had a mean of 10.7, SD 3.0 extractions and the placebo group had a mean preparacetamol HOPS score of 7.5, SD 3.0 out of a maximum of 14. We feel confident that sufficient pain was present to demonstrate the analgesic properties of the test medicine.
Measuring paediatric patient's pain and the relief thereof, is extremely difficult [1,23]. Methods include self assessment, behavioural and physiological measurements. In this study, physiological measurements showed no differences between the groups and as these comprise three (blood pressure, pulse, respiratory rate) of the seven categories of the HOPS, we had to rely on the behavioural categories (crying, movement, agitation, verbalization) to provide a score. Nevertheless, the HOPS is a validated form of paediatric pain assessment [2,16]. Self assessment is the most reliable form of pain measurement and the OFPS is well documented for use in children older than 3 years of age, below the entry level of the present study [23,24].
The HOPS, Table 2, does not show a difference in the pain experienced by the two groups, however, it shows the overall pain experienced with and without additional post-operative paracetamol. It must be assessed in the light of the 82.8% children in the placebo group who received paracetamol compared with the 19.4% in the tramadol group (P<0.01). Reinforcing the demonstration of analgesic efficacy is the OFPS, Figs. 1 and 2. Particularly Fig. 2, in which the paracetamol effect has been removed, shows the improved analgesia on awakening. In adults, the maximum serum concentration following oral tramadol is reached at 2 h , so the analgesia is likely to improve during the immediate recovery phase. This seems to be the case as Fig. 2 shows the severity of the tramadol group's pain to be half that of the placebo groups pain for the first hour and one third for the second hour.
Knight  found that parents consistently scored higher pain levels than nurses, and Goodenough et al. reported that nurses and children's assessments were similar.
Interestingly, our maternal assessments were almost identical with those of the children and both were similar to those of the nursing staff. Cultural factors may play a role here, the patients coming from a stoical socio-economic group.
In conclusion, we have demonstrated that the addition of oral tramadol 1.5 mg kg−1 to oral midazolam 0.5 mg kg−1 in children aged 4-7 years, has no deleterious effects on the behavioural or physiological responses during the peri-operative period. The analgesia provided is effective and reliable for the discomfort due to multiple dental extractions. The overall duration of action was in excess of 3 h.
The project was carried out with support from Grünenthal GmbH, Stolberg, Germany. The authors are gratefully indebted to Dr S. Ratcliffe for valuable assistance throughout the study.
1 Mather L, Mackie J. The incidence of post-operative pain in children. Pain
2 Hannallah RS, Broadman LM, Belman AB, Abramowitz MD, Epstein BS. Comparison of candal and ilioinguinal/iliohypogastric nerve blocks for the control of post-orchidopexy pain in paediatric ambulatory surgery. Anaesthesiology
3 Littlejohn IH, Tarling MM, Flynn AJ, Ordman AJ, Aiken A. Post-operative pain relief in children following extraction of carious deciduous teeth under general anaesthesia: a comparision of nalbuphine and diclofenac. Euro J Anaesth
4 Knight JC. Post-operative pain in children after day case surgery. Paediatric Anaesthesia
5 Corman HH, Harnick EJ, Kritchman M, Forestman N, Terestman N. Emotional reactions of surgical patients to hospitalisation, anaesthesia and surgery. Am J Surg
6 Acs G, Moore PA, Shusterman S, Needleman HL. The extent of trauma and postextraction pain in children. Pediatric Dentistry
7 Sunshine A. New clinical experience with tramadol. Drugs
(Suppl. 1): 8-18.
8 Dayer P. The pharmacology of tramadol. Drugs
(Suppl. 1): 3-7.
9 Schäffer J. Investigation of tramadol for post-operative analgesia in children. Fortschritte Anästhesiologie
10 Barsoum MN. Comparison of the efficacy and tolerability of tramadol, pethidine and nalbuphine in children with post-operative pain. Clin Drug Invest
11 Roelofse JA, Van der Bijl P, Stegmann DH, Hartshorne JE. Pre-anaesthetic medication with rectal midazolam in children undergoing dental extraction. J Oral Maxillofac Surg
12 Feld LH, Negus JB, White PF. Oral midazolam pre-anesthetic medication in pediatric outpatients. Anesthesiology
13 Yaster M, Nichols DG, Deshpande JK. Midazolam-fentanyl intravenous sedation in children: case report of respiratory arrest. Pediatrics
14 Ramsay MAE, Savage TM, Simpson BRJ, Goodwin R. Controlled sedation with alphaxalone-alphadolone. Br Med J
15 Aldrete JA, Kroulik D. A post-anaesthetic recovery score. Anesth Analg
16 Norden J, Hannallah R, Getson P, O'Donell R, Kelliher G, Walker N. Reliability of an objective pain scale in children. Anesth Analg
17 Fletcher B. Psychological upset in posthospitalized children - a review. Matern Child Nurse J
18 Padfield NL, Twohig M, Fraser AC. Temazepam and trimeprazine compared with placebo as premedication in children. Br J Anaes
19 Bevan JC, Johanston C, Tousignant E, Kiron V, Carranga R. Pre-operative parenteral anxiety predicts behavioural and emotional responses to induction of anaesthesia in children. Can Anaesth Soc J
20 Moore PA, Acs G, Hargreaves JA. Postextraction pain relief in children: a clinical trial of liquid analgesics. Int J Clin Pharmacol Ther Toxicol
21 McGraw T. Analgesics in pediatric dentistry, relative efficacy of aluminum ibuprofen suspension and acetaminophen elixir. ASDC J Dent Child
22 Payne KA, Hendrix MRG, Wade WJ. Caudal bupivacaine for post-operative analgesia in paediatric lower limb surgery. J Pediat Surg
23 Morton NS. Pain assessment in children. Paediatric Anaesthesia
24 Goodenough B, Addicoat L, Champion GD et al.
Pain in 4-to 6-year-old children receiving intramuscular injections: a comparison of the faces pain scale with other selfreport and behavioral measures. The Clin J Pain