Pain Medicine

Section Editor: Spencer S. Liu

Pain Management After Elective Hallux Valgus Surgery: A Prospective Randomized Double-Blind Study Comparing Etoricoxib and Tramadol

Metha Brattwall, MD,* Ibrahim Turan, MD, PhD,† and Jan Jakobsson‡

BACKGROUND: Pain is a common complaint after day surgery, and there is still a controversy surrounding the use of selective cyclooxygenase-2 (COX-2) inhibitors. In the present prospective, randomized, double-blind study we compared with a selective (COX-2) inhibitor (etoricoxib) with pain management using sustained-release tramadol after elective hallux valgus surgery. METHODS: One hundred ASA 1 to 2 female patients were randomized into 2 groups of 50 patients each; oral etoricoxib 120 mg ϫ 1 ϫ IV ϩ 90 mg ϫ 1 ϫ day V–VII and oral tramadol sustained-release 100 mg ϫ 2 ϫ VII. Pain, pain relief, satisfaction with pain management, and need for rescue medication were evaluated during the first 7 postoperative days. A computed tomography scan evaluating bone healing was performed 12 weeks after surgery. A clinical evaluation of outcome (healing, mobility, and patient-assessed satisfaction) was performed 16 weeks after surgery. RESULTS: Two patients withdrew before discharge from the hospital. Ninety-eight patients, 81 ASA 1 and 17 ASA 2 (82 nonsmokers and 14 smokers), mean age 49 years (19–65), weight 64 (47–83) kg, and height 167 (154–183) cm were evaluated. Overall pain was well managed, but the mean visual analog scale (VAS) was significantly lower among etoricoxib patients evaluated during the entire 7-day period studied (12.5 Ϯ 8.3 vs. 17.3 Ϯ 11, P Ͻ 0.05). patient’s grading of pain relief (92 Ϯ 12 vs. 85 Ϯ 15, P Ͻ 0.05) and satisfaction with pain medication (47/49 vs. 39/49, P Ͻ 0.05) was higher among etoricoxib patients. Patients receiving tramadol reported significantly more side effects. Six patients, all in the tramadol group, discontinued the study because of side effects (P Ͻ 0.05). At 14-day follow-up 1 patient in the etoricoxib group and 5 patients in the tramadol group exhibited minor irritation in the wound area. The 12-week computed tomography scan showed good healing in 82 patients, 43 in the etoricoxib group, and 39 in the tramadol group. The study found ongoing healing in 11 patients, 4 in the etoricoxib group and 7 in the tramadol group. The 16-week patient-assessed Health Profile Quality of life revealed high patient satisfaction overall; 47 patients in each study group rated the outcome as satisfactory and the mean change in the patient-assessed quality of life VAS score was 6.2 and 2.6 for the etoricoxib and tramadol groups, respectively. Clinical follow-up at 16 weeks showed high functionality and no signs or symptoms of improper healing in any patient. CONCLUSION: Etoricoxib was found to be more effective and associated with fewer side effects in comparison with tramadol sustained release as a component of multimodel analgesia after elective hallux valgus surgery. There were no signs of impaired wound or bone healing associated with the use of etoricoxib. (Anesth Analg 2010;111:544–9)

ain and nausea are common complaints after day case facilitate resumption of the normal activities of daily living.3 orthopedic surgery. Simple and effective pain medica- Nonsteroidal anti-inflammatory drugs (NSAIDs) have been tion after discharge is of importance to maintain quality shown to be effective analgesics in the early postoperative P 1,2 of care. The goal in optimizing postoperative pain manage- phase.4–7 The selective cyclooxygenase-2 (COX-2) inhibitors ment is to reduce pain, improve the quality of recovery, and may provide perioperative analgesia without the impaired hemostasis associated with nonselective NSAIDs. For ex- From the *Department of Anaesthesia, Sahlgrenska Academy, University ample, Desjardins et al. found 50 mg significantly Hospital Mo¨ndal, Gothenburg, Sweden; †Foot & Ankle Surgical Centre, Stock- holm, Sweden; ‡Department of Anaesthesia & Intensive Care, Karolinska more effective than was diclofenac sodium 100 mg for total Institutet, Institution for Physiology & Pharmacology, Stockholm, Sweden. pain relief, onset time, and duration of response after Accepted for publication March 30, 2010. bunionectomy.8 The institution received a grant from Merck Sharp & Dome. The potential negative effects from long-term NSAID Jan Jakobsson is currently affiliated with Foot & Ankle Surgical Centre, use on the healing process have always been controversial.9 Stockholm, Sweden. Conflict of Interest: The study had financial support by a grant from Merck Studies involving postoperative administration of COX-2 Sharp & Dome selective inhibitors for 10 to 14 days after cardiac surgery Address correspondence and reprint requests to Jan Jakobsson, Professor, demonstrated an increased risk of postoperative wound Foot & Ankle Surgical Centre, Stora¨ngsva¨gen 10, 11542 Stockholm, Sweden. infection and cardiovascular complications.10,11 This has Address e-mail to [email protected]. Copyright © 2010 International Anesthesia Research Society led physicians to question the wisdom of using COX-2 12 DOI: 10.1213/ANE.0b013e3181e3d87c selective inhibitors after orthopedic surgery, which was

544 www.anesthesia-analgesia.org August 2010 • Volume 111 • Number 2 compounded by the withdrawal of several COX-2 selective spontaneously through a facemask. Administration of inhibitors,13 and the retractions of Scott Reuben’s studies sevoflurane was stopped when skin suturing began. Surgery on NSAIDs.a was performed with a standardized technique including an A recent meta-analysis from the Cochrane library sup- osteotomy of the first metatarsal bone, excision of the exosto- ports the utility of a single dose of etoricoxib 120 mg for sis, and a split fixation of the osteotomy to allow full weight postoperative pain in low-risk patients. One study suggests bearing immediately after surgery.17 All procedures were that a 120-mg etoricoxib dose is as effective as, or better performed by a senior orthopedic surgeon (Ibrahim Turan) than, other commonly used analgesics.14 The effects of with Ͼ25 years of experience in forefoot surgery. etoricoxib for pain management for a longer period of time All patients were fast-tracked to the step-down unit, (e.g., up to 1 week after surgery) has not been extensively where they were provided the initial dose of study medi- studied. We have observed in noncontrolled studies the cation in combination with 30 mg paracetamol per kilo- analgesic efficacy of etoricoxib after ligament repair and gram orally. bunion surgery.15,16 Before discharge a dedicated research nurse instructed The aim of the present prospective, randomized, all patients on how to take study and rescue medications, if double-blind study was to compare the analgesic efficacy of needed, and how to complete the daily diary during the etoricoxib to that of tramadol sustained release as a com- first 7 days after surgery. Study medication and rescue ponent of multimodal analgesia during the first 7 days after analgesics were provided at discharge. All patients were elective hallux valgus surgery. provided oral paracetamol1gasfirst-line rescue and oral oxycodone 5 mg as second-line rescue analgesia and in- METHODS structed how to take rescue analgesia in an escalating The study protocol was approved by the local ethical manner, when needed. committee (2008/828 to 319) and EudraCT registered (Eu- A dedicated research nurse phoned each patient 1 to 2 draCT 2008 to 000,791–24) before inclusion of any patients. days after discharge to reinforce adherence to the study After receiving written informed consent, 100 female ASA protocol. All patients were seen by a research nurse 12 to 16 1 to 2 patients were randomized into 2 groups by computer days after surgery to assess wound healing and remove randomization: stitches. A computed tomography (CT) scan with multiplanar (a) etoricoxib 120 mg oral once daily for 4 days followed reformatted images along the axis of the first metatarsal was by 90 mg oral once daily for 3 days; conducted at 12 weeks to assess bone healing. The CT scans (b) tramadol sustained release 100 mg twice daily for 7 were evaluated by 2 physicians blinded to the randomization days. and otherwise not involved in the study. The images were The first dose was provided immediately after the end of assessed on a 3-grade scale: well-established healing, ongoing surgery when patients were adequately awake in the bone healing, and limited bone healing. recovery area. The study medication was prepared by a The Euro Quality 5 Dimension (EQ-5D) questionnaire hospital pharmacist in identical-appearing capsules accord- was used for a patient-assessed health profile quality of b ing to a computer-generated random number schedule. The life at baseline before surgery and at 16 weeks in conjunction patients, nurses, surgeons, and anesthesiologists directly with an outpatient clinical assessment of healing and func- involved in patient care were blinded to the content of the tionality by the patient and a physician otherwise not in- oral study medication capsules. volved in the study. The EQ-5D questionnaire is a generic tool Patients were provided oral and written pain medica- commonly used for assessing health profile quality of life. It tion information, and instructed on how to complete pain addresses the self-assessed experience of pain, mobility, activ- questionnaires (visual analog scale [VAS] 0 to 10, with 10 ity, self-care, and mode/depression with a pregraded scale defined as the worst possible pain), medication pain relief (impaired or yes/some impairment or some symptoms/no ϭ questionnaire (VAS scale 0 to 100, with 100 defined as symptoms adequate function). It also consists of a quality of ϭ ϭ perfect pain relief), satisfaction with pain medication (bi- life VAS scale, 100 excellent quality of life and 0 death. nary, satisfied/unsatisfied), and presence of any side ef- Functionality was assessed by a defined protocol with pr- fects (asked as an open question). No adverse symptoms egraded scales: satisfied, neutral, or dissatisfied. were explicitly asked for, and no scale was used for grading of side effects. Statistics All patients followed the same perioperative protocol. All data are presented as mean and SD unless stated Patients fasted after midnight on the night before surgery. otherwise. Data were analyzed using analysis of variance Upon arrival in the operating room, IV access was estab- (ANOVA) and ␹2 analysis as appropriate. Continuous lished, and 8 mg IV betamethasone was given. Anesthesia variables were evaluated using ANOVA. When multiple was induced with 3 to 3.5 ␮g/kg alfentanil and 2.5 mg/kg comparisons of continuous data (e.g., pain scores) were propofol. Ten milliliters 1% lidocaine was injected in the performed between the treatment groups, a Bonferroni wound area before washing and dressing. Anesthesia was correction was applied. Categorical variables were evalu- maintained with sevoflurane in an oxygen–air mixture. ated using ␹2 contingency table analysis. For all statistical Sevoflurane was titrated to maintain heart rate and arterial analyses, P values Ͻ0.05 were considered to be statistically blood pressure within 15% of baseline. Patients breathed significant after correction for multiple comparisons. a See http://www.aaeditor.org/HWP/Retraction.Notice.pdf. b See http://www.euroqol.org/.

August 2010 • Volume 111 • Number 2 www.anesthesia-analgesia.org 545 Etoricoxib Versus Tramadol for Postoperative Pain

Table 1. Patients’ Demographics Etoricoxib Tramadol P (49 ؍ n) (49 ؍ n) Age (years) 47 Ϯ 11 51 Ϯ 10 ns Weight (kg) 62 Ϯ 866Ϯ 9 ns Length (cm) 167 Ϯ 6 168 Ϯ 5 ns Smokers 7 7 ns ASA 1–2 41/8 40/9 ns Quality of life VAS score prior to 88 Ϯ 21 92 Ϯ 19 ns surgery from the EQ-5D (1–100) Statistical analysis: analysis of variance and ␹2 contingency table analysis. ns, nonsignificant difference between study groups; VAS ϭ visual analog scale; EQ-5D, Euro Quality 5 Dimension questionnaire.

Table 2. Intraoperative Observations Etoricoxib Tramadol P (49 ؍ n) (49 ؍ n) Duration of surgery 11.4 Ϯ 1.2 11.4 Ϯ 1.2 ns (minutes) Duration of anaesthesia 12.7 Ϯ 1.5 12.6 Ϯ 1.5 ns (minutes) Propofol (mg) 155 Ϯ 33 158 Ϯ 21 ns Alfentanil (mg) 0.21 Ϯ 0.04 0.23 Ϯ 0.04 0.0408 Mean EtSevo (%) 1.1 Ϯ 0.3 1.2 Ϯ 0.3 ns Pain VAS 30 minutes 1.9 Ϯ 1.0 2.2 Ϯ 2.0 ns Discharge eligible 55 Ϯ 456Ϯ 4 ns (minutes) Statistical analysis: analysis of variance and ␹2 contingency table analysis. ns, nonsignificant difference between study groups; EtSevo ϭ end tidal sevoflurane; VAS ϭ visual analog scale.

Table 3. Results—Follow-up at Day 14 Etoricoxib Tramadol P (49 ؍ n) (49 ؍ n) Compliance (no. of patients) Completed pain therapy 49 43 0.035 Discontinued 0 6 Rescue analgesics (No. of patients) No rescue 20 13 ns Figure 1. Consort flow chart of patients. Paracetamol 18 22 Oxycodon 11 14 Side effects (No. of patients.) No side effects 41 14 Ͻ0.01 The number of patients studied was estimated from a Emesis 4 18 power analysis based on earlier findings of the experience Dizziness 0 9 of pain and need for rescue analgesics after hallux valgus Other gastrointestinal 2 2 surgery. A group size of 48 was chosen to have a 90% Sleepiness 2 4 power to detect a difference with a P Ͻ 0.05 for the number Nose bleed 0 1 Infection 0 1 of patients requiring rescue opioid analgesic. To compen- Brief pain inventory score (1–70) sate for an estimated dropout of 2 patients in each group, Day 3 20 Ϯ 17 23 Ϯ 16 100 patients were randomized. Day 7 15 Ϯ 15 16 Ϯ 14 Satisfied with pain management yes/no (no. of patients) 47/2 39/10 0.031 RESULTS Wound irritation at 14 days 15ns (no. of patients) One hundred two female patients were screened, and 100 were enrolled in the study. All randomized patients were in Statistical analysis: analysis of variance and ␹2 contingency table analysis. good health. No patient had therapy-dependent chronic ns, nonsignificant difference between study groups. Brief pain inventory: the impact from pain during the last day on 7 domains: pain. Two patients withdrew on the day of surgery. A activity, mode, ambulation, work, social relations, sleep, and quality of life CONSORT flow chart is provided in Figure 1. Ninety-eight rated on visual analogue scales 0–10, no influence ϭ 0, extensively ϭ patients received study medication. Table 1 summarizes influenced 10. the demographics and preoperative quality-of-life scores for patients receiving study medication. No major complications were reported during the Surgery and anesthesia were uneventful. All patients 2-week follow-up period (Table 3). Six patients discontin- were discharged within 2 hours after the end of surgery ued study medication because of side effects: nausea, (Table 2). dizziness, and sleepiness. During the examination on day

546 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA Table 4. Patients’ Assessed Outcome at 16-week Follow-up Etoricoxib Tramadol P (49 ؍ n) (49 ؍ n) Patient’s assessed outcome Fast walking 0/6/43 2/7/38 ns Passive mobility 3/11/30 4/7/34 ns Active mobility 3/13/33 6/4/37 ns Healing process 0/6/43 3/7/39 ns Swollen 0/15/34 2/8/39 ns Satisfaction with surgical outcome 0/4/45 1/4/44 ns Blinded physician assessed outcome Passive mobility 2/7/36 1/6/42 ns Active mobility 2/8/29 1/6/41 ns Healing process 0/5/37 2/6/39 ns Swollen 0/10/33 0/4/44 ns Satisfaction with surgical outcome 0/2/45 1/4/44 ns No. of patients: impaired/neutral/good. Statistical analysis: ␹2 contingency table analysis. ns, nonsignificant difference between study groups.

Table 5. Patients’ Assessed Health Profile Quality of Life (EQ5D) at 16-week Follow-up Etoricoxib Tramadol P (49 ؍ n) (49 ؍ Figure 2. The upper graph in the figure shows pain relief from study (n medication for the etoricoxib and tramadol groups of patients (mean Mobility 0/6/43 0/6/43 ns and SD). The lower graph in the figure shows daily maximum pain Pain visual analog scale (VAS) scores for the etoricoxib and tramadol Pain 1/7/41 0/11/37 ns groups of patients (mean and SD). “Discomfort” 0/12/37 0/13/36 Activities 14, we found minor surgical site irritation and incomplete Work 0/2/47 0/0/48 ns Family 0/0/49 0/0/49 ns wound closure in 6 patients: 5 active controls (tramadol) Leisure 0/6/42 0/9/39 ns and 1 in the etoricoxib group. Self-care 0/0/49 0/0/49 ns Thirty-three patients, 20 in the etoricoxib group and 13 Mode in the tramadol group, did not take any rescue medication Depressed/anxiety 0/2/46 0/3/46 ns Satisfied with surgery 0/1/47 0/2/47 ns during the 7-day follow-up period. Twenty-five patients Quality of life VAS at 16 weeks 94 ͓60– 100͔ 95 ͓60– 100͔ ns took at least 1 dose of oxycodone as a rescue analgesic. Difference in quality of life VAS, There was no significant difference between the groups after–before 6.2 Ϯ 21 2.6 Ϯ 20 ns (Table 3) in the use of oxycodone. No. of patients: impaired or yes/some impairment or some/no symptoms— Forty patients reported side effects. Side effects (pre- adequate function. dominantly nausea, dizziness, and sleepiness) were re- Statistical analysis: analysis of variance and ␹2 contingency table analysis. ϭ ported more frequently by the tramadol group (Table 3). VAS visual analog scale. One patient in the tramadol group had minor bleeding from the nose. No cardiovascular side effects or symptoms the tramadol group. None of the CT scans showed were reported. Satisfaction with pain management was limited bone healing. There was no association between rated significantly higher among the etoricoxib patients smoking, age, or body weight and grading of the CT (P Ͻ 0.05) (Table 3). image for bone repair. The 11 patients graded as showing VAS pain scores were generally low. The maximum ongoing bone healing reported more pain in their health daily pain VAS grading showed a distinct pattern with a profile questionnaire at 16 weeks than did those with peak on days 2 and 3 after surgery and a reduction well-established healing: 3/11 vs. 14/82 among the 93 thereafter. The mean maximum VAS was significantly patients assessed by CT scan. lower among etoricoxib patients evaluated during the At the 16-week follow-up visit there was satisfactory entire 7-day period studied (12.5 Ϯ 8.3 vs. 17.3 Ϯ 11 P Ͻ healing and functionality in all patients (Table 4 A, B). No 0.05). A significant difference in daily maximum pain VAS signs of impaired bone or wound healing were observed. scores was seen on days 3, 4, and 7 (P Ͻ 0.05) (Fig. 2). Pain All patients scored healing, mobility, and satisfaction with relief from study medication was rated as high for patients surgery as high (Table 4A). The blinded physician assess- in both groups, but was significantly higher for the etori- ment at the 16-week follow-up visit did not reveal any coxib group (P Ͻ 0.05) on days 2, 3, and 5 (Fig. 2). between-groups differences (Table 4B). The patient- Ninety-three patients had a CT scan, showing a gener- assessed quality-of-life health profile, EQ-5D, showed mod- ally reassuring healing process. Eighty-two patients had est improvement and no differences between groups well-established healing, 43 in the etoricoxib group and 39 (Table 5). The patients’ self-assessed quality-of-life VAS in the tramadol group. Eleven patients were graded as increased by 6.2 in the etoricoxib group and 2.6 in the ongoing bone healing, 4 in the etoricoxib group and 7 in tramadol group, respectively. Satisfaction with surgical

August 2010 • Volume 111 • Number 2 www.anesthesia-analgesia.org 547 Etoricoxib Versus Tramadol for Postoperative Pain outcome was high, with 47 patients in each group assessing wound healing was observed from the administration of the outcome as satisfactory (Table 5). ketorolac.27 The effect of nonselective NSAIDs on bone healing was also analyzed in a retrospective study in DISCUSSION children who had major spine surgery. No signs of im- NSAIDs are frequently used in day case surgery as part of paired bone healing were seen in this retrospective analy- 28 multimodal pain management to provide adequate pain sis. A 2-year follow-up study of patients who received relief and to reduce the risk of opioid-related side effects.18 postoperatively for 14 days after total knee re- Hallux valgus surgery has been suggested as a useful placement found no signs of increased risk of loosening of model for testing multiple-day analgesic therapy. It has the prosthesis, and concluded that celecoxib may be used 29 also been used for studying the effects of opioid analgesics safely in conjunction with total knee replacement. Be- as well as NSAIDS.8,19,20 We found a structured 7-day pain cause we studied ASA 1 to 2 patients only, and the dose of analgesic protocol to be associated with acceptable pain etoricoxib was reduced for the last 3 days, our findings levels and high patient satisfaction after elective hallux might not apply to patients given larger doses, patients valgus surgery. The etoricoxib group had lower maximum given doses for a longer period of time, or patients who pain scores during the entire 7-day period and significantly have other risk factors for poor healing such as smoking, better analgesic effects on the most painful days (days 2 corticosteroids use, diabetes, or delayed healing or non- 12 and 3 after surgery). Etoricoxib was also associated with union of a fracture. fewer side effects and thus overall higher patient satisfac- We used the EQ-5D to assess patients’ evaluation of and tion with pain medication. In our view, the most important satisfaction with outcome, an accepted tool for evaluation 20 observation is that we found no differences in the healing of outcomes in health care. There were no differences process. Both the clinical assessment and the CT scan at 12 between the groups at 16 weeks. weeks demonstrated equivalent healing of the wound and Our findings are limited to healthy, ASA 1 to 2 patients. the bone in both groups. The cardiovascular and thrombotic risks associated with 30,31 We could not verify our primary outcome variable, a NSAIDs will be higher in a less healthy population. We difference in need for rescue medication. However, pain studied only female patients, who are more prone to 32,33 VAS scores were lower and satisfaction with pain medica- postoperative pain, nausea, and vomiting. We studied tion was higher in the etoricoxib group. Toivonen et al. a peripheral orthopedic procedure, not a major procedure compared the effects of a single 120- mg dose of etoricoxib such as a joint replacement or spine surgery. Additionally, given 1 hour before placebo control in patients undergoing all patients were provided a multimodal analgesia, includ- arthroscopic acromioplasty performed under regional an- ing the use of local anesthesia infiltration before incision, IV esthesia. They found beneficial effects on pain and a betamethasone, oral paracetamol upon arrival in the step- reduced need for rescue analgesia during the entire 7-day down unit, and perioperative opioids. Lastly, we used an study period from a single dose.21 Sun et al. studied the active control rather than a placebo control. This was done effects of celecoxib 200 mg twice daily for 3 days after for both ethical reasons, and because tramadol is well plastic surgery, and found less need for opioid analgesics established as an effective analgesic. than for placebo during the 3-day study.22 We did not We conclude that etoricoxib 120 mg oral once daily for 4 include a placebo, but an active control, tramadol slow days followed by 90 mg oral once daily for 3 days provides release twice daily. Tramadol is a well-known centrally better analgesia than does tramadol sustained-release 100 mg acting analgesic with an analgesic potency of about 10% of twice daily for 7 days. We also conclude that in ASA 1 to 2 morphine but with less effect on ventilation.23,24 Tramadol patients this regimen of postoperative etoricoxib is not asso- has been shown to be effective for the treatment of ortho- ciated with deleterious effects on the healing process. pedic postoperative pain.25 We chose tramadol sustained release in a fixed dose of 100 mg twice daily as a ACKNOWLEDGMENTS comparison. While this may have been a low dose, We would like to thank Aleris Diagnostics and Dr. Tengvar for resulting in less analgesia when compared with the support with CT imaging and evaluation of CT scans. etoricoxib dose, it was associated with a higher incidence REFERENCES of side effects. We could increase the analgesic effect by 1. Shnaider I, Chung F. Outcomes in day surgery. Curr Opin giving the subjects a larger dose of tramadol, but would Anaesthesiol 2006;19:622–9 have seen increased complications. 2. Janssen KJ, Kalkman CJ, Grobbee DE, Bonsel GJ, Moons KG, Vergouwe Y. The risk of severe postoperative pain: modifica- We did not see any clinical or radiological signs of tion and validation of a clinical prediction rule. Anesth Analg impaired healing, but our findings should be put into 2008;107:1330–9 perspective. Evaluation of the bone-healing process is not 3. White PF, Kehlet H. Postoperative pain management and easy. Modeling of the CT image may improve the accuracy patient outcome: time to return to work! Anesth Analg of evaluation,26 which is why we used multiplanar refor- 2007;104:487–9 4. Barden J, Derry S, McQuay HJ, Moore RA. Single dose oral matted CT images to improve evaluation of nonunion of ketoprofen and dexketoprofen for acute postoperative pain in fractures. However, our study cannot absolutely exclude an adults. Cochrane Database Syst Rev 2009;4:CD007355 increase of clinical pseudoarthrosis from impaired wound 5. Derry C, Derry S, Moore RA, McQuay HJ. Single dose oral healing. ibuprofen for acute postoperative pain in adults. Cochrane Database Syst Rev 2009;3:CD001548 There is a growing body of evidence suggesting that the 6. Lloyd R, Derry S, Moore RA, McQuay HJ. Intravenous or effects of NSAIDs on bone repair are of limited clinical intramuscular parecoxib for acute postoperative pain in adults. importance. In a recent animal study no impairment in Cochrane Database Syst Rev 2009;2:CD004771

548 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA 7. Derry P, Derry S, Moore RA, McQuay HJ. Single dose oral 21. Toivonen J, Pitko VM, Rosenberg PH. Etoricoxib pre- diclofenac for acute postoperative pain in adults. Cochrane medication combined with intra-operative subacromial block Database Syst Rev 2009;2:CD004768 for pain after arthroscopic acromioplasty. Acta Anaesthesiol 8. Desjardins PJ, Black PM, Daniels S, Bird SR, Fitzgerald BJ, Scand 2007;51:316–21 Petruschke RA, Tershakovec A, Chang DJ. A randomized 22. Sun T, Sacan O, White PF, Coleman J, Rohrich RJ, Kenkel JM. controlled study comparing rofecoxib, diclofenac sodium, and Perioperative versus postoperative celecoxib on patient out- placebo in post-bunionectomy pain. Curr Med Res Opin comes after major plastic surgery procedures. Anesth Analg 2004;20:1523–37 2008;106:950–8 9. Haws MJ, Kucan JO, Roth AC, Suchy H, Brown RE. The effects 23. Grond S, Sablotzki A. Clinical pharmacology of tramadol. Clin of chronic ketorolac tromethamine (toradol) on wound heal- Pharmacokinet 2004;43:879–923 ing. Ann Plast Surg 1996;37:147–51 10. Ott E, Nussmeier NA, Duke PC, Feneck RO, Alston RP, Snabes 24. Scott LJ, Perry CM. Tramadol: a review of its use in perioper- MC, Hubbard RC, Hsu PH, Saidman LJ, Mangano DT. Efficacy ative pain. Drugs 2000;60:139–76 and safety of the cyclooxygenase 2 inhibitors parecoxib and 25. Bourne MH, Rosenthal NR, Xiang J, Jordan D, Kamin M. in patients undergoing coronary artery bypass Tramadol/acetaminophen tablets in the treatment of postsur- surgery. J Thorac Cardiovasc Surg 2003;125:1481–92 gical orthopedic pain. Am J Orthop 2005;34:592–7 11. Nussmeier NA, Whelton AA, Brown MT, Langford RM, Hoeft 26. Fayad LM, Patra A, Fishman EK. Value of 3D CT in defining A, Parlow JL, Boyce SW, Verburg KM. Complications of the skeletal complications of orthopedic hardware in the postop- COX-2 inhibitors parecoxib and valdecoxib after cardiac sur- erative patient. AJR Am J Roentgenol 2009;193:1155–63 gery. N Engl J Med 2005;352:1081–91 27. Eck JC, Gomez BA, Yaszemski MJ. The effects of postoperative 12. Boursinos LA, Karachalios T, Poultsides L, Malizos KN. Do ketorolac on wound healing in a rat model. Orthopedics steroids, conventional non-steroidal anti-inflammatory drugs 2009;32:827 and selective Cox-2 inhibitors adversely affect fracture heal- 28. Sucato DJ, Lovejoy JF, Agrawal S, Elerson E, Nelson T, ing? J Musculoskelet Neuronal Interact 2009;9:44–52 McClung A. Postoperative ketorolac does not predispose to 13. Alacqua M, Trifiro` G, Cavagna L, Caporali R, Montecucco CM, pseudoarthrosis following posterior spinal fusion and instru- Moretti S, Tari DU, Galdo M, Caputi AP, Arcoraci V. Prescrib- mentation for adolescent idiopathic scoliosis. Spine (Phila Pa ing pattern of drugs in the treatment of osteoarthritis in Italian 1976) 2008;33:1119–24 general practice: the effect of rofecoxib withdrawal. Arthritis 29. Meunier A, Aspenberg P, Good L. Celecoxib does not appear Rheum 2008;59:568–74 to affect prosthesis fixation in total knee replacement: a ran- 14. Clarke R, Derry S, Moore RA, McQuay HJ. Single dose oral etoricoxib for acute postoperative pain in adults. Cochrane domized study using radiostereometry in 50 patients. Acta Database Syst Rev 2009;2:CD004309 Orthop 2009;80:46–50 15. Turan I, Assareh A, Rolf C, Jakobsson JG. Etoricoxib, paraceta- 30. Varas-Lorenzo C, Castellsague J, Stang MR, Perez-Gutthann S, mol, and dextropropoxyphene for postoperative pain manage- Aguado J, Rodriguez LA. The use of selective cyclooxygenase-2 ment: a questionnaire survey of consumption of take-home inhibitors and the risk of acute myocardial infarction in medication after elective hallux valgus surgery. Foot & Ankle Saskatchewan, Canada. Pharmacoepidemiol Drug Saf 2009; Specialist 2008;1:88–92 18:1016–25 16. Ekenman I, Brattwall M, Turan I, Jakobsson J. Quality audit of 31. Chan CC, Reid CM, Aw TJ, Liew D, Haas SJ, Krum H. Do an office-based “fast track” elective ankle ligament repair COX-2 inhibitors raise blood pressure more than nonselective program. The Foot and Ankle Online J 2009;2:4–8 NSAIDs and placebo? An updated meta-analysis. J Hypertens 17. Lindgren U, Turan I. A new operation for hallux valgus. Clin 2009;27:2332–41 Orthop Relat Res 1983;175:179–83 32. Gan TJ, Meyer TA, Apfel CC, Chung F, Davis PJ, Habib AS, 18. Allen SC, Ravindran D. Perioperative use of nonsteroidal Hooper VD, Kovac AL, Kranke P, Myles P, Philip BK, Samsa G, anti-inflammatory drugs: results of a UK regional audit. Clin Sessler DI, Temo J, Trame`r MR, Vander Kolk C, Watcha M, Drug Investig 2009;29:703–11 Society for Ambulatory Anesthesia. Society for Ambulatory 19. Apfelbaum JL, Desjardins PJ, Brown MT, Verburg KM. Anesthesia guidelines for the management of postoperative Multiple-day efficacy of parecoxib sodium treatment in post- nausea and vomiting. Anesth Analg 2007;105:1615–28 operative bunionectomy pain. Clin J Pain 2008;24:784–92 33. Mei W, Seeling M, Franck M, Radtke F, Brantner B, Wernecke 20. Daniels SE, Upmalis D, Okamoto A, Lange C, Ha¨eussler J. A randomized, double-blind, phase III study comparing multiple KD, Spies C. Independent risk factors for postoperative pain in doses of tapentadol IR, oxycodone IR, and placebo for postop- need of intervention early after awakening from general an- erative (bunionectomy) pain. Curr Med Res Opin 2009;[Epub aesthesia. Eur J Pain 2009;[Epub ahead of print] ahead of print]

August 2010 • Volume 111 • Number 2 www.anesthesia-analgesia.org 549