Double-blind, randomized, placebo-controlled trial comparing rofecoxib with dexketoprofen trometamol in surgical dentistry

I. D. Jackson*,1, B. H. Heidemann2, J. Wilson2, I. Power2 and R. D. Brown1

1 Edinburgh Dental Institute, Lauriston Building, Lauriston Place, Edinburgh EH3 9YW, UK. 2 University Department Critical Care and Pain Medicine, The Royal Infirmary, Little France, Edinburgh EH16 9SU, UK

*Corresponding author. E-mail: iainandjulia{at}waitrose.com

Accepted for publication: December 5, 2003.


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background. Rofecoxib, a selective cyclooxygenase-2 inhibitor, and dexketoprofen trometamol, a single isomer non-steroidal anti-inflammatory drug (NSAID), are available for the treatment of acute pain. Both are claimed to have fewer adverse effects than traditional NSAIDs. We have compared them in a clinical setting.

Methods. We performed a double-blind randomized controlled trial involving 120 patients undergoing surgical removal of a single mandibular third molar at the Edinburgh Dental Institute. Those who developed moderate pain within 4 h of the procedure were allocated to one of three groups: rofecoxib 50 mg (Group RO, n=37); dexketoprofen trometamol 25 mg (Group DE, n=42); or placebo (Group PL, n=41). Participants monitored pain intensity and pain relief for 24 h using visual analogue scales (VAS) and verbal rating scales (VRS). The summed, time-weighted pain relief score to 8 h derived from the VRS (TOTPAR 8) was used as the primary outcome variable.

Results. No significant difference was demonstrated between Groups RO and DE using TOTPAR 8 as the primary outcome variable. Both drugs were significantly different compared with placebo. Rescue analgesia during the trial period was required by only 15 out of 37 subjects in Group RO, but 35 out of 42 subjects in Group DE. The median times to use of rescue medication were 150 (Group PL), 398 (Group DE) and 1440 min (Group RO). Both drugs were well tolerated and adverse events reported were mild to moderate in severity.

Conclusions. Rofecoxib and dexketoprofen trometamol are effective treatments for acute pain using a dental pain model and are well tolerated. Rofecoxib has a longer duration of action as a single dose and gave adequate analgesia for over half of that study group; patients in the dexketoprofen trometamol group needed more rescue analgesia.

Br J Anaesth 2004; 92: 675–80

Keywords: anaesthesia, dental; pain; pharmacology, COX-2 inhibitors


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
This study compared dexketoprofen trometamol, a non-selective non-steroidal anti-inflammatory drug (NSAID) and single isomer (S(+)-enantiomer) of ketoprofen formulated as a trometamol salt,1 with rofecoxib, a selective cyclooxygenase-2 (COX-2) inhibitor,2 using the post-dental surgery model. Dexketoprofen trometamol has been demonstrated to be effective in the treatment of acute pain.35 COX-2 inhibitors such as rofecoxib claim to match the analgesic efficacy of non-specific NSAIDs69 whilst producing fewer adverse effects (e.g. gastric ulceration).1013 With the exception of one study in rats and mice,14 rofecoxib and dexketoprofen trometamol have not yet been compared.

Rofecoxib has the theoretical advantage of being a once daily formulation and could therefore improve patient compliance and satisfaction in this setting. If proven to be efficacious, further research could explore if preoperative use of rofecoxib results in the avoidance of moderate or severe pain.15 This study aims to compare the efficacy of both drugs in a placebo-controlled trial.

The use of the post-dental surgery pain model involving the surgical removal of lower third molar teeth for the development, assessment and comparison of analgesic drugs is well described in the literature. It is particularly suitable for this area of research as the majority of subjects are young, healthy adults, the degree of postoperative pain experienced is consistent and at least of moderate nature in the majority of patients.1618


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Approval for this double-blind randomized controlled trial was granted by Lothian Regional Ethics Committee and written informed patient consent was obtained. The study was conducted in the oral surgery department of the Edinburgh Dental Institute between June 2001 and July 2002.

Routine consultation clinics were used to screen all patients referred for surgical removal of a lower third molar tooth for eligibility and willingness to participate in the trial. Inclusion criteria were: age 18–60 yr, extractions requiring bone removal; extractions possible under local anaesthesia; development of pain of at least moderate nature as described by a four-point verbal rating scale (VRS, 1=absent, 2=mild, 3=moderate, 4=severe); and request for analgesia within 4 h of the completion of surgery.

Exclusion criteria were: any past medical or drug history contraindicating the use of NSAIDs (e.g. allergy, risk of drug interactions, gastrointestinal disease, asthma); pregnancy; breast-feeding; current signs and symptoms of acute infection; use of analgesics within 12 h of surgery; and postoperative complications, e.g. patient feeling unwell.

The principle author performed 96% of extractions. A standardized quantity of local anaesthetic (lidocaine 2%, 3.3 ml with epinephrine 1:80 000) was administered for each procedure and where necessary the block was supplemented with further increments of the same solution. After surgery, all patients remained in the oral surgery department until the development of moderate pain. Four subjects failed to develop moderate pain within the required time span. Those developing moderate pain were administered a single dose of rofecoxib 50 mg (Group RO), dexketoprofen trometamol 25 mg (Group DE), or placebo (Group PL), each dose representing the optimal therapeutic dose previously described.4 19

Patients were advised that should their level of pain relief be insufficient at any time during the 24 h trial period, they could make use of rescue analgesia (codeine 30 mg with acetaminophen 500 mg, provided to all subjects). Subjects were requested to avoid rescue analgesia where possible in the first 60 min after the administration of trial medication to allow time for the trial medication to take effect.

After administration of study medication, patients returned home with instructions to undertake pain intensity and pain relief assessments using both 100 mm visual analogue scales (VAS) and VRS (pain intensity recorded as above; pain relief 1=none, 2=a little, 3=some, 4=a lot, 5=complete). From this, point assessments were recorded in a diary by the patient every 15 min for the first hour, and then hourly for 7 h, with a final assessment at 24 h. The diary was returned to the investigators in a stamped and addressed envelope.

Global evaluations of pain relief were completed at 8 h and 24 h, and before the use of rescue medication using a five-point VRS (1=poor to 5=excellent). The time of onset of first meaningful pain relief and of first use of rescue medication were recorded. The number of patients taking rescue medication during the study period and within the proposed duration of action of the active drugs was recorded. Adverse events experienced were written into the diary by subjects as free text.

Randomization was carried out in the pharmacy department of the Royal Infirmary of Edinburgh. All study drugs were presented in identical capsules with patient numbers only on the packaging. Coding was not broken until the end of the study.

The summed, time-weighted pain relief to 8 h using data from a VRS, TOTPAR 8 (out of 32), was used as the primary outcome variable. Based on this, a power calculation was performed (Minitab v.12) before commencement of the trial. A 4 point or greater difference in the primary outcome variable was regarded as clinically significant and using previously published data we assumed the mean for Group PL to be 10 with a standard deviation of 6. Setting a significance level of P=0.05 it was predicted that a group size of 35 patients would allow detection of a difference between groups with a power of 80%.

All data were collated in a custom-made data-entry application (Microsoft Access 2000) with VAS readings being recorded to the nearest millimetre and VRS readings being assigned numerical values.

The following scores were calculated as indices of drug efficacy:4 6 the summed, time-weighted pain relief to 8 h using both VAS and VRS data [TOTPAR 8 (VRS) and TOTPAR 8 (VAS)]; the summed, time-weighted pain intensity difference to 8 h using both VAS data (SAPID 8) and VRS data (SPID 8). TOTPAR 8 data derived from both VRS and VAS were calculated as follows:

Pain relief (PARt) = PARbaseline–PARtime t

TOTPAR 8 = {Sigma} PARtxtime (h) elapsed since previous observation.

Related data were obtained from the pain intensity scores, namely the summed, time weighted pain intensity difference to 8 h. SAPID and SPID scores were calculated as follows:

Pain intensity difference (PIDt) = PIbaseline–PItime t

SAPID/SPID 8 = {Sigma} PIDtxtime (h) since previous observation.

The greater the value of each of the summed, time weighted pain relief and pain intensity difference scores, the greater the pain relief and the less pain intensity experienced, respectively.

The data were analysed on a PC running Microsoft Windows 2000 Professional. Statistical analysis and graphical representation were performed using Minitab (v.12) and Graphpad Prism (v.3). Statistical analysis was based on the intention-to-treat analysis.20

Values for the primary end-point of TOTPAR 8 (VRS) and the secondary endpoints of TOTPAR 8 (VAS), SPID 8 and SAPID 8 were compared using one-way analysis of variance (ANOVA). Differences between groups were then assessed using Bonferroni’s correction for multiple comparisons. Patient characteristics were analysed using ANOVA, Bonferroni’s correction for multiple comparisons and {chi}2-tests as appropriate. The data for first meaningful pain relief (FMPR) and time to first use of rescue medication are described as medians, and the differences between the two active drugs further analysed using ANOVA. Global evaluation of pain relief at 8 h, 24 h and before the use of rescue medication is given as median (interquartile range). The data for the 8 and 24 h time points were analysed using ANOVA and Hsu’s multiple comparisons with the best test. The data at the time of taking rescue medication were analysed using ANOVA and Bonferroni’s correction for multiple comparisons.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
In total, 127 patients were recruited and after obtaining written informed consent had surgical extraction of a lower third molar. In 123 patients (96.9%) the procedure led to pain of at least a moderate nature and these were randomized to receive placebo (Group PL, n=43), dexketoprofen trometamol 25 mg (Group DE, n=42), or rofecoxib 50 mg (Group RO, n=38). Of those randomized, 120 concluded the trial. No data were available for two patients in the placebo group, one questionnaire was lost in the post and one patient failed to send in his questionnaire despite repeated attempts at contact. One patient in Group RO was not dosed because of severe nausea. The patient characteristics are given in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient characteristics. Data are presented as mean (range) for age, or mean (SD) unless indicated. No statistically significant differences observed between groups. The {chi}2-test was used on sex; ANOVA on age, BMI, and alcohol consumption
 
Four patients in Group PL failed to complete a sufficient number of pain relief records to calculate the TOTPAR 8, whilst all patients in Groups DE and RO completed all assessments.

There was no statistically significant differences in the primary end-point variable TOTPAR 8 (VRS) between Groups RO and DE (Table 2). Both were, however, statistically different to Group PL with P<0.001 for Group RO vs PL and P<0.01 for Group DE vs PL.


View this table:
[in this window]
[in a new window]
 
Table 2 Pain relief and pain intensity difference data. Data are presented as mean (95% CI), or median (interquartile range). TOTPAR, summed, time weighted pain relief score; SPID, summed, time weighted pain intensity difference score from VRS; SAPID, summed, time weighted pain intensity difference score from VAS. Statistical significance tests were by ANOVA followed by Bonferroni’s correction for multiple comparisons. *P<0.01 vs Group PL; **P<0.05 vs Group PL; ***P<0.001 vs Group PL; {dagger}P<0.05 vs Group DE; {ddagger}P=0.6 vs Group DE
 
The data for pain relief (PAR) by treatment group over the initial 8 h of the trial is presented in Figure 21. This time–effect graph gives an indication of the mean values of PAR at each time interval.



View larger version (13K):
[in this window]
[in a new window]
 
Fig 2 Time course of raw pain relief scores.

 


View larger version (20K):
[in this window]
[in a new window]
 
Fig 1 Flow of participants.

 
The values for TOTPAR (VAS) and SAPID/SPID at 8 h after medication derived from visual analogue and VRS readings are presented in Table 2.

Using TOTPAR 8 (VAS) and SPID 8, no statistical difference was found between the two active drugs. However, SAPID 8 showed a statistically significant difference between Groups RO and DE (P<0.05).

The median (interquartile range) times obtained for FMPR in each group are presented in Table 2. The median time to FMPR for Group RO was 89 min compared with 60 min in Group DE. This difference was statistically non-significant (ANOVA, P=0.6). As the value for FMPR in Group PL represents the onset of pain relief from rescue analgesia it was not compared with the treatment groups.

The median (interquartile range) for global evaluation of pain relief at 8 and 24 h after medication as well as immediately before the first intake of rescue analgesia are summarized in Table 3. Global evaluation of pain relief at 8 and 24 h shows a significant difference for both active drugs against placebo. At 24 h global pain relief is also rated as significantly better with rofecoxib compared with dexketoprofen trometamol using Hsu’s MCB test.


View this table:
[in this window]
[in a new window]
 
Table 3 Global pain relief evaluation and rescue medication data. Data are presented as median (interquartile range). Global pain relief at 8 and 24 h for Group DE, and at 8 h for Group RO was significantly different from Group PL using Hsu’s MCB test. Global pain relief at 24 h for Group RO was significantly different from Group PL and Group DE using Hsu’s MCB test. *P<0.001 vs Group PL (using ANOVA and Bonferroni’s correction for multiple comparisons); {dagger}P<0.001 vs Group DE ({chi}2-test); {ddagger}P=0.00009 vs Group DE
 
Data regarding the use of rescue analgesia in the treatment groups are also recorded in Table 3. Only 15 patients in Group RO required rescue analgesia during the 24 h trial period compared with 35 patients in Group DE and 36 patients in Group PL. These data were compared using a {chi}2-test, which showed a significant difference (P<0.001) between the groups.

The median times to the first use of rescue medication were 150, 398 and 1440 min for Groups PL, DE and RO, respectively. There is a statistically significant difference between Groups DE and RO using ANOVA with P=0.0009.

Fifteen of the 120 subjects described 20 adverse events after the administration of study medication, six in Group RO, five in Group DE, and four in Group PL. The most frequent events were nausea and vomiting (n=4), tiredness/drowsiness (n=3), headache (n=2), and feeling hot/feverish (n=2). One patient in Group DE reported profuse bleeding. No serious adverse events were reported and none of the reported events resulted in withdrawal from the study.


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
NSAIDs and COX-2 inhibitors are capable of providing good pain relief in situations of moderate to severe pain. Therefore, in order to provide a worthwhile test of the analgesic capabilities of rofecoxib and dexketoprofen trometamol it was necessary to undertake a surgical procedure capable of generating such levels of pain. By undertaking bone removal in a surgical extraction it is highly likely that the patient will develop pain of at least a moderate nature.21 22 In this study, 97% of patients developed moderate pain within the required 4 h time period after surgery, confirming the usefulness of this model in producing pain of a predictable intensity.

No statistically significant difference was shown between the two active components in our primary outcome variable TOTPAR 8 (VRS). However, rofecoxib achieved higher values for all summed time weighted pain relief and pain intensity scores and in the case of SAPID 8 this achieved statistical significance.

Rarely in dental pain studies is one analgesic agent shown to display a significant difference in efficacy relative to another. Whilst SAPID 8 showed a significant difference between Groups RO and DE (P<0.05; 95% CI Group DE 81.9–163.8; 95% CI Group RO 172.8–268.8) and no significant difference between Groups DE and PL (95% CI Group PL 13.6–125.1), this finding was not reproduced by the primary outcome variable TOTPAR 8 (VRS), nor other secondary variables (i.e. TOTPAR 8 (VAS) and SPID 8).

A potential reason for not being able to show a difference between the two active groups lies in the fact that local anaesthetics such as lidocaine have been shown to decrease prostaglandin and eicosanoid production. This could, at least theoretically, lead to the augmentation of one or both of the active drugs.1 23

Our data confirm the manufacturers claims that dexketoprofen trometamol provides pain relief within 30 minutes and rofecoxib within 45 min. The difference seen in our data is, however, not statistically significant.

Peak levels of pain relief for both dexketoprofen trometamol and rofecoxib were not reached until the second hour after drug administration. This has implications for the timing of analgesia administration relative to dental surgery and preoperative use of analgesia in the dental pain model could be an area of interest for future research. Analgesia could be administered before or on completion of surgery. The former option has already demonstrated promise with rofecoxib 50 mg being administered 1 h before spinal fusion surgery.24

The PAR-time graph (Fig. 2) shows a gradual and relatively slow increase in the first hour. A much greater increase is seen between the 1 h and 2 h assessments. At this time peak pain relief is reached for dexketoprofen trometamol whereas rofecoxib stays just below its peak. This suggests that more frequent assessments up to the 2 h stage might have been useful.

Our data show that 25 patients in Group DE required rescue medication before the claimed duration of action for dexketoprofen trometamol. Fifteen patients in Group RO required rescue medication within the claimed duration of action of rofecoxib. The time to the first use of rescue medication was statistically significantly different between the two active drugs. Despite the use of rescue medication in Group DE, pain relief is rated better at the end of the study period in Group RO. These findings suggest that rofecoxib provides better overall analgesia over a 24 h period and might improve patient satisfaction and compliance.

Whilst it has been described that VAS exhibits greater sensitivity than VRS,25 many authors believe it easier for subjects to relate to a VRS. The VRS based TOTPAR has been used as a primary outcome variable for several previous studies.46 8 9 11 12 For easier comparison with previous studies, we have used TOTPAR 8 based on a VRS as the primary outcome variable. VRS and VAS for pain relief and pain intensity were also recorded.

In order to generate a TOTPAR 8 score, previous studies recorded all scores from the use of rescue medication as severe. This approach does not take into account that, for some individuals, the natural time course of the pain experienced would have led to a decrease in their pain scores in the period after rescue analgesia use. McGurk4 used a TOTPAR over 6 h when investigating dexketoprofen trometamol as its claimed duration of action is only 8 h. However, to compare a long-acting analgesic with a short-acting analgesic where the onset of action is claimed faster in the short acting drug a 6 h TOTPAR might not be appropriate.

Using an intention-to-treat analysis, both of these problems can be overcome. The cost of this approach, where pain scores are recorded even after rescue medication has been given, is that any difference between the groups becomes smaller. However, any results gained become more significant. Applying the intention-to-treat principle allows unbiased assessment of the efficacy of treatments at the level of protocol adherence observed in the trial. This level of adherence is likely to be closely related to that in the community and thus reflects the effectiveness of the treatments that can be expected in the community.20 Intention-to-treat analysis also minimizes the influence of undetected confounding factors.20

As the recommended therapeutic regimen for dexketoprofen trometamol is 25 mg three times daily, meaningful comparison between dexketoprofen trometamol 25 mg and rofecoxib 50 mg could not be carried beyond the 8 h point. A single 24 h assessment was included in order to test the claim of the manufacturers that rofecoxib 50 mg produces relief of acute pain over a 24 h period.

The incidence of adverse events was well within expected limits.3 4 6 7 This appears to support the claims that rofecoxib and dexketoprofen trometamol are well tolerated, owing to the COX-2 specificity of the former and the single enantiomer formulation of the latter.

One of the adverse events noted was ‘profuse bleeding’ commencing 4 h after the administration of dexketoprofen trometamol and lasting for 30 min. Although previous trials of this nature have dismissed bleeding as an event more related to the surgical procedure rather than any medication taken, it should be remembered that dexketoprofen trometamol does have an inhibitory effect on platelet aggregation.

Our data show rofecoxib 50 mg and dexketoprofen trometamol 25 mg to be efficient analgesics in this pain model. No statistically significant difference was found between rofecoxib and dexketoprofen tometamol with regards to the primary outcome variable (TOTPAR 8 (VRS)). The time to onset was not significantly different for the two drugs. A single dose of dexketoprofen trometamol or rofecoxib provided sufficient analgesia for the complete 24-hour study period in 40% and 60% of patients respectively. Rofecoxib provided better overall pain relief at the 24-hour time point. Both drugs were equally well tolerated with adverse events being mild to moderate in severity.


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1 Mauleon D, Artigas R, Garcia ML, Carganico G. Preclinical and clinical development of dexketoprofen. Drugs 1996; 52 (Suppl. 5): 24–45[ISI][Medline]

2 Ehrich EW, Dallob A, De Lepeleire I, et al. Characterization of rofecoxib as a cyclooxygenase-2 isoform inhibitor and demonstration of analgesia in the dental pain model. Clin Pharmacol Ther 1999; 65: 336–47[ISI][Medline]

3 Bagan JV, Lopez Arranz JS, Valencia E, et al. Clinical comparison of dexketoprofen trometamol and dipyrone in postoperative dental pain. J Clin Pharmacol 1998; 38 (Suppl. 12): 55S–64S[Abstract/Free Full Text]

4 McGurk M, Robinson P, Rajayogeswaran V, et al. Clinical comparison of dexketoprofen trometamol, ketoprofen, and placebo in postoperative dental pain. J Clin Pharmacol 1998; 38 (Suppl. 12): 46S–54S[Abstract/Free Full Text]

5 Gay C, Planas E, Donado M, et al. Analgesic efficacy of low doses of dexketoprofen in the dental pain model. A randomised, double-blind, placebo-controlled study. Clin Drug Invest 1996; 11: 320–30[ISI]

6 Morrison BW, Christensen S, Yuan W, Brown J, Amlani S, Seidenberg B. Analgesic efficacy of the cyclooxygenase-2-specific inhibitor rofecoxib in post-dental surgery pain: a randomized, controlled trial. Clin Ther 1999; 21: 943–53[CrossRef][ISI][Medline]

7 Malmstrom K, Daniels S, Kotey P, Seidenberg BC, Desjardins PJ. Comparison of rofecoxib and celecoxib, two cyclooxygenase-2 inhibitors, in postoperative dental pain: a randomized, placebo- and active-comparator-controlled clinical trial. Clin Ther 1999; 21: 1653–63[CrossRef][ISI][Medline]

8 Scott LJ, Lamb HM. Rofecoxib. Drugs 1999; 58: 499–505[ISI][Medline]

9 Jeske AH. COX-2 inhibitors and dental pain control. J Gt Houst Dent Soc 1999; 71: 39–40[Medline]

10 Hunt RH, Bowen B, Mortensen ER, et al. A randomized trial measuring fecal blood loss after treatment with rofecoxib, ibuprofen, or placebo in healthy subjects. Am J Med 2000; 109: 201–6[CrossRef][ISI][Medline]

11 Lanza FL, Rack MF, Simon TJ, et al. Specific inhibition of cyclooxygenase-2 with MK-0966 is associated with less gastroduodenal damage than either aspirin or ibuprofen. Aliment Pharmacol Ther 1999; 13: 761–7[CrossRef][ISI][Medline]

12 Simon LS, Lanza FL, Lipsky PE, et al. Preliminary study of the safety and efficacy of SC-58635, a novel cyclooxygenase 2 inhibitor: efficacy and safety in two placebo-controlled trials in osteoarthritis and rheumatoid arthritis, and studies of gastrointestinal and platelet effects. Arthritis Rheum 1998; 41: 1591–602[ISI][Medline]

13 Bjarnason I, Sigthorsson G, Crane R, et al. COX-2 specific inhibition with MK-0966 25 or 50 mg q.d. does not increase intestinal permeability: a controlled study with placebo (PBO) and indomethacin 50 mg t.i.d. (INDO). Am J Gastroenterol 1998; 93: 246

14 Mazario J, Gaitan G, Herrero JF. Cyclooxygenase-1 vs cyclooxygenase-2 inhibitors in the induction of antinociception in rodent withdrawal reflexes. Neuropharmacology 2001; 40: 937–46[CrossRef][ISI][Medline]

15 Reuben SS, Bhopatkar S, Maciolek H, Joshi W, Sklar J. The preemptive analgesic effect of rofecoxib after ambulatory arthroscopic knee surgery. Anesth Anal 2002; 94: 55–9[Abstract/Free Full Text]

16 Urquhart E. Analgesic agents and strategies in the dental pain model. J Dent 1994; 22: 336–41[CrossRef][ISI][Medline]

17 Cooper SA. Models for clinical assessment of oral analgesics. Am J Med 1983; 75: 24–9[ISI][Medline]

18 Cooper SA, Beaver WT. A model to evaluate mild analgesics in oral surgery outpatients. Clin Pharmacol Ther 1976; 20: 241–50[ISI][Medline]

19 Morrison BW, Fricke J, Brown J, Yuan W, Kotey P, Mehlisch D. The optimal analgesic dose of rofecoxib: overview of six randomized controlled trials. J Am Dent Assoc 2000; 131: 1729–37[ISI][Medline]

20 Montori VM, Guyatt GH. Intention-to-treat principle. CMAJ 2001; 165: 1339–41[Free Full Text]

21 Seymour RA, Blair GS, Wyatt FA. Post-operative dental pain and analgesic efficacy. Part I. Br J Oral Surg 1983; 21: 290–7[ISI][Medline]

22 Seymour RA, Blair GS, Wyatt FA. Post-operative dental pain and analgesic efficacy. Part II. Analgesic usage and efficacy after dental surgery. Br J Oral Surg 1983; 21: 298–303[ISI][Medline]

23 Jonsson A, Cassuto J, Tarnow P, Sinclair R, Bennett A, Tavares IA. Effects of amide local anaesthetics on eicosanoid formation in burned skin. Acta Anaesthesiol Scand 1999; 43: 618–22[CrossRef][ISI][Medline]

24 Reuben SS, Connelly NR. Postoperative analgesic effects of celecoxib or rofecoxib after spinal fusion surgery. Anesth Analg 2000; 91: 1221–5[Abstract/Free Full Text]

25 Seymour RA. The use of pain scales in assessing the efficacy of analgesics in post-operative dental pain. Eur J Clin Pharmacol 1982; 23: 441–4[ISI][Medline]