Acta Anaesthesiol Scand 2014; ••: ••–•• Printed in Singapore. All rights reserved

© 2014 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd ACTA ANAESTHESIOLOGICA SCANDINAVICA

doi: 10.1111/aas.12380

Review Article

Adverse effects of perioperative paracetamol, NSAIDs, glucocorticoids, gabapentinoids and their combinations: a topical review O. Mathiesen1, J. Wetterslev2, V. K. Kontinen3, H.-C. Pommergaard4, L. Nikolajsen5, J. Rosenberg4, M. S. Hansen6, K. Hamunen7, J. J. Kjer8, J. B. Dahl6 and Scandinavian Postoperative Pain Alliance (ScaPAlli) 1 Section of Acute Pain Management, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark, 2Copenhagen Trial Unit, Centre for Clinical Intervention Research, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark, 3Department of Anaesthesia and Intensive Care, Helsinki University Central Hospital, Helsinki, Finland, 4Department of Surgery, Herlev Hospital, University of Copenhagen, Herlev, Denmark, 5Department of Anaesthesiology, Aarhus University Hospital, Aarhus, Denmark, 6Department of Anaesthesiology, Centre of Head and Orthopaedics, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark, 7Pain Clinic, Helsinki University Central Hospital, Helsinki, Finland and 8Department of Gynecology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark

Post-operative pain affects millions of patients worldwide and the post-operative period has high rates of morbidity and mortality. Some of this morbidity may be related to analgesics. The aim of this review was to provide an update of current knowledge of adverse events (AE) associated with the most common perioperative non-opioid analgesics: paracetamol, non-steroidal anti-inflammatory drugs (NSAIDs), glucocorticoids (GCCs), gabapentinoids and their combinations. The review is based on data from systematic reviews with meta-analyses of analgesic efficacy and/or adverse effects of perioperative non-opioid analgesics, and randomised trials and cohort/retrospective studies. Generally, data on AE are sparse and related to the immediate post-operative period. For paracetamol, the incidence of AEs appears trivial. Data are inconclusive regarding an association of NSAIDs with mortality, cardiovascular events, surgical bleeding and renal impairment. Anastomotic leakage may be associated with NSAID usage. No firm evidence exists for an association of

P

ost-operative pain affects millions of patients worldwide1 and the post-operative period has high rates of morbidity and mortality.2–6 It is well established that effective perioperative pain management with minimal adverse effects will facilitate early mobilisation, fluid and food intake and the resumption of normal physical activities, which is crucial for the surgical patient’s well-being and optimal rehabilitation after surgery.7–9 In contemporary pain management, patients are most often treated with a combination of different non-opioid analgesics, in order to reduce opioid requirements and the well-known adverse effects

NSAIDs with impaired bone healing. Single-dose GCCs were not significantly related to increased infection rates or delayed wound healing. Gabapentinoid treatment was associated with increased sedation, dizziness and visual disturbances, but the clinical relevance needs clarification. Importantly, data on AEs of combinations of the above analgesics are sparse and inconclusive. Despite the potential adverse events associated with the most commonly applied non-opioid analgesics, including their combinations, reporting of such events is sparse and confined to the immediate perioperative period. Knowledge of benefit and harm related to multimodal pain treatment is deficient and needs clarification in large trials with prolonged observation. Accepted for publication 04 July 2014 © 2014 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

associated with opioids.7–9 Many assortments of non-opioid analgesics are currently used in clinical practice. However, our knowledge about the potential adverse events (AEs) related to such combinations is insufficient. In a previous update10 we addressed the potential benefits of paracetamol, non-steroidal antiinflammatory drugs (NSAIDs), selective cyclooxygenase-2 (COX-2) inhibitors, glucocorticoids (GCCs), gabapentin, pregabalin and their combinations, with regard to pain-relieving and opioidsparing effects, and opioid-related AEs in surgical patients.10

1 bs_bs_banner

O. Mathiesen et al.

In the present topical review, we aim to update the knowledge of specific AEs associated with these drugs, including their combinations, in postoperative pain intervention.

adverse effects were not reported in a number of meta-analyses considering multiple dosing with oral or i.v. paracetamol for up to 24 h postoperatively13–16 (Table 1).

Literature search

Data from other studies. Recently, concerns have been raised about the COX-2 inhibition by paracetamol in endothelium,17 which may explain the cardiovascular risk associated with the longterm use in epidemiological studies. Furthermore, epidemiological data suggest an increased gastrointestinal risk of adverse effects associated with high-dose paracetamol when given for a prolonged time or when co-administered with traditional NSAIDs.17 Likewise, a recent large randomised trial investigating a treatment for 13 weeks with paracetamol, ibuprofen or the combination hereof for knee pain in general practice patients, discovered similar falls in haemoglobin levels for paracetamol and ibuprofen, most likely caused by occult gastrointestinal bleeding.18

Due to the broad scope of this topical review, neither the literature search nor the data provided may be exhaustive. The review is primarily based on data from previous systematic reviews with metaanalyses of analgesic efficacy and/or adverse effects of non-opioid analgesics, and randomised clinical trials (RCTs), cohort studies and retrospective studies, where appropriate. MEDLINE, EMBASE and the Cochrane Central Register of Controlled trials were searched for systematic reviews, randomised clinical trials and cohort/retrospective studies. Free-text combinations including the following search terms were used: paracetamol, acetaminophen, NSAID, coxib, COX-2 inhibitor, cyclooxygenase-2 inhibitor, steroid, dexamethasone, glucocorticoids, gabapentin, pregabalin, pain, post-operative, adverse effects. Reference lists of relevant articles were checked to identify additional relevant studies. The last search was performed on 01 April 2014. Inclusion criteria. We included systematic reviews with meta-analyses, RCTs, cohort studies and retrospective studies, investigating analgesic efficacy and drug-specific AEs of non-opioid analgesics after surgery. Studies in adults, who had undergone minor to major surgical procedures, receiving systemic paracetamol, NSAIDs, COX-2 inhibitors, GCCs, gabapentinoids or a combination of these analgesics, were included in the review. Metaanalyses that predominantly included drugs that are no longer licensed (valdecoxib and rofecoxib) were excluded from the review and so were analyses addressing local application of analgesics.

Results Adverse effects of individual drugs Paracetamol Data from systematic reviews and meta-analyses of analgesic efficacy. The incidence of adverse effects with a single dose of paracetamol is very low and in most studies not different from that of the placebo group,11,12 except for pain on infusion with intravenous (i.v.) propacetamol12 (Table 1). However,

2

NSAIDs and selective COX-2 inhibitors Data from systematic reviews and meta-analyses of analgesic efficacy. The analysis by Moore et al.19 summarised data from 35 Cochrane reviews, including 46 drug/dose combinations with the majority of the original studies performed in patients undergoing dental surgery. All included reviews assessed the effect of single doses of oral analgesics taken alone. Thirty-seven of the drug/dose combinations included conventional NSAIDs or selective COX-2 inhibitors, of which nine were assessed in acute postoperative pain other than dental surgery (including episiotomy, orthopaedic and abdominal surgery). The analyses included results for participants experiencing ‘at least one adverse event’. Such events were few and not different between NSAIDs/COX-2 inhibitors and placebo, except for aspirin 600/ 650 mg [numbers needed to harm (NNH) 44] and diflunisal 1000 mg (NNH 7.7).19 It is acknowledged by the authors, however, that studies using a single dose of analgesic represent a poor test of AEs. Several meta-analyses addressing analgesic efficacy have also reported on different serious AEs, including all-cause mortality and cardiovascular events,20 surgical bleeding,14,15,20–22 GI-bleeding14,15,20 and renal impairment.14,15,20 Results from these analyses are summarised in Table 2. Note that most of these meta-analyses include at least partly the same original trials, which should be taken into the consideration in the overall interpretation.

Adverse effects of multimodal analgesia Table 1 Systematic reviews with meta-analyses of perioperative paracetamol/propacetamol in post-operative pain treatment: reported adverse effects. References

Drug and administration N (active/placebo)

Surgical population

Adverse effects

Toms et al. 200811

Oral paracetamol, single dose 3277/2425 (51 trials)

Not statistically different from placebo. RR 1.12 (95% CI 0.97, 1.29)

Tzortzopoulou et al. 201112

i.v. paracetamol or propacetamol, single dose. A total of 3896 patients were analysed (36 trials)

Oral surgery; episiotomy; caesarean section; abdominal, gynaecologic, urological or orthopaedic surgery; tonsillectomy ‘Any kind’ of surgery

Remy et al. 200513

Oral and i.v. paracetamol/propacetamol; multiple dosing/24 h 265/226 (7 trials) Oral and i.v. paracetamol/propacetamol; multiple dosing/24 h 379/334 (10 trials) Oral and i.v. paracetamol/propacetamol, multiple dosing/24 h. Trials in adults that reported 24 h cumulative dose of morphine delivered by a PCA device were considered N not provided (12 trials) i.v. paracetamol. Included studies had to report PON, PONV or post-operative vomiting, either as a primary end point, secondary end point, or as adverse events 1223/1141 (30 trials)

Elia et al. 200514 McDaid et al. 201015

Apfel et al. 201316

Mixed, major surgery

Not different from placebo; patients receiving i.v. propacetamol had more pain on infusion than placebo and i.v. paracetamol No data

Mixed, major surgery

No data

Mixed, major surgery

None reported

Mixed, minor and major surgery

No data

CI, confidence interval; i.v., intravenous; PCA, patient-controlled analgesia; PON, post-operative nausea; PONV, post-operative nausea and vomiting; RR, relative risk.

Duration of NSAID treatment ranged from a single dose (one trial) up to multiple doses during 24 h post-operatively (three trials) and was not reported in two trials. Follow-up periods were from 48 h post-operatively (two trials), up to 30 days postoperatively (one trial), or duration of follow-up was not reported (three trials). Incidence of surgical bleeding was significantly increased in two of five analyses reporting this outcome.14,21 Number of patients with severe surgical bleeding also increased from 0/259 (0%) in controls to 7/410 (1.7%) with NSAIDs (NNH = 59).14 No significant differences from placebo could be demonstrated with regard to cardiovascular events,20 GI-bleeding14,15,20 or renal failure (Table 2).14,15,20 Data from other studies. A number of meta-analyses, RCTs as well as cohort studies and retrospective

studies have investigated specific AEs related to perioperative treatment with NSAIDs and/or selective COX-2 inhibitors.23–30 Duration of NSAID/ COX-2 treatment varied from 2 to 12 days. Follow-up periods were from 24 h to 30 days postoperatively (Table 3). Overall, the results could not demonstrate an effect of NSAIDs on overall mortality, cardiovascular events, surgical bleeding or renal impairment in patients without kidney disease and with normal pre-operative kidney function (Table 3). It is noteworthy that the overall number of fatalities was very low in two studies including 11.245 patients undergoing mixed major surgery (0.2%),25 and 10.873 patients undergoing total hip or knee joint replacement (0%),26 respectively. The post-operative follow-up periods in these studies were 30 days25 and 115 h,26 respectively. The low numbers of

3

4 Mixed surgery (not dental)

Trials of a single-dose perioperative systemic ketorolac with an inactive (placebo or ‘no treatment’) control group that reported pain scores or opioid consumption as post-operative outcomes in subjects older than 18 years

n = 400/389 (17 trials)

Randomised trials comparing the use of NSAIDs/COX-2 inhibitors in addition to opioid analgesics vs. opioid analgesics alone

A total of 1065 patients were analysed (20 trials)

Randomised allocation to a NSAID-containing single- or multiple-dose analgesic regimen vs. non-NSAID-containing opioid or regional analgesic regimen given pre-, intra- or post-operatively in adult patients

A total of 782 patients were analysed (13 trials)

N not specified, 16 trials

Lumbar discectomy, laminectomy, spinal fusion

Cardiac or thoracic surgery

Mixed, major surgery

N not specified (38 trials)

Oral or i.v. COX-2 inhibitors; single dose, multiple dose. Trials in adults that reported 24 h cumulative dose of morphine delivered by a PCA device were considered

Not reported

Not reported

Single dose

Up to 24 h

Up to 24 h

Up to 24 h

Duration of NSAID treatment

No data

All-cause mortality at 30 days: 0% vs. 5%, NS (OR 0.19, 95% CI 0.01, 4.22), NS

No data

No data

No data

No data

Death, NSAID vs. control

No data

Myocardial infarction: 1.1% vs. 1.5% (NS) Atrial fibrillation: 10.1% vs. 14.3% (NS)

No data

No data

No data

No data

Cardiovascular events, NSAID vs. control

NS (numbers not reported)

Bleeding, all causes: 1.1% vs. 1% (OR = 0.72, 95% CI 0.09, 5.66), NS

Abnormal bleeding significantly increased with ketorolac: OR = 2.43 (95% CI 0.5, 11), P < 0.05 (?)

Not different from placebo (few data)

10/422 (2.4%) vs. 1/273 (0.4%), NS

13/760 (1.7%) vs. 1/604 (0.2%), OR = 4.54 (95% CI 1.54, 13.4), P < 0.05, NNH = 65

Surgical bleeding, NSAID vs. control

No data

No data

No data

No data

No data

7/410 (1.7%) vs. 0/259 (0%), OR 6.08 (95% CI 1.33, 27.9), P < 0.05, NNH = 59

Severe surgical bleeding, NSAID vs. control

No data

1.3% vs. 1.5% (OR 0.96, 95% CI 0.13, 7.09), NS

No data

0/100 vs. 0/100

4/171 (2.3%) vs. 0/211 (0%), NS

4/171 (2.3%) vs. 0/111, OR = 5.12 (95% CI 0.65, 40.6), NS

GI-bleeding, NSAID vs. control

No data

4.9% vs. 5.5% (OR 0.95, 95% CI 0.37, 2.46), NS

No data

No data

1/111 (0.9%) vs. 0/105, OR = 7.03 (95% CI 0.14, 355), NS

Renal failure, NSAID vs. control

No data

No data

No data

No data

25/580 (4.3%) vs. 9/429 (2.1%), NS

24/560 (4.3) vs. 9/409 (2.2%), OR = 1.69 (95% CI 0.82, 3.47), NS

Oliguria NSAID vs. control

No data

No data

No data

No data

No data

No data

Anastomotic leakage, NSAID vs. control

Not reported

Up to 30 days post-operatively, but generally not specified. Actual numbers not provided

Not reported

Up to 48 h post-operatively

Up to 48 h post-operatively

Not reported

Follow-up period

GI, gastrointestinal; IM, intramuscular; i.v., intravenous; NSAID, non-steroidal anti-inflammatory drug; OR, odds ratio; CI, confidence interval; NNH, numbers needed to harm; NS, not significant; PCA, patient-controlled analgesia; COX-2, cyclooxygenase-2.

Jirarattanaphochai and Jung 200822

Bainbridge et al. 200620

De Oliveira et al. 201221

McDaid et al. 201015 (Cox)

Mixed major surgery

Oral, rectal, intranasal, IM or i.v. non-selective NSAID; single dose, multiple-dose, or continuous infusions. Trials in adults that reported 24 h cumulative dose of morphine delivered by a PCA device were considered

n = 1304/1147 (33 trials)

Mixed major surgery

Oral, IM or i.v. NSAID; single dose, multiple-dose or continuous infusions. Trials in adults that reported the 24 h cumulative dose of morphine delivered by a PCA device were considered

Elia et al. 200514

McDaid, et al. 201015 (NSAID)

Surgical procedure

Drug and administration N (active/placebo)

References

Systematic reviews with meta-analyses of perioperative NSAIDs/COX-2 inhibitors in post-operative pain treatment: reported adverse effects.

Table 2

O. Mathiesen et al.

Burton et al. 201327

At least 1 dose of a NSAID was given perioperatively within 48 hours of surgery

Mean duration 3 days

Hip or knee replacement

Liu et al. 201226

Surgery with formation of at least 1 anastomosis of the small bowel, colon or rectum

Up to 7 days

Mixed major surgery

Forrest et al. 200225

Meta-analysis of RCTs comparing an NSAID or COX-2 inhibitor with a control treatment in adults (age > 16 years) undergoing surgery with formation of at least 1 anastomosis of the small bowel, colon or rectum. Only trials reported in 1990 or later were included; 272 patients received NSAID or COX-2 inhibitor, 208 were controls, 6 trials

Up to 12 days

Mixed moderate to major surgery

Merry et al. 200424

Not accurately stated

Mixed, elective surgery

Meta-analysis of randomised or quasi-randomised comparisons of individual NSAIDs and COX-2 inhibitors with either each other or placebo for treatment of post-operative pain, with relevant post-operative renal outcome measures, in adult surgical patients with normal renal function; n = 1459, 23 trials RCT of tenoxicam vs. placebo; 750 received tenoxicam, 251 patients received placebo RCT of ketorolac vs. diclofenac or ketoprofen; 11245 patients received NSAID, no placebo group Observational cohort study. 9221 patients received NSAID (meloxicam or ketorolac); 610 patients received COX-2inhibitor (celecoxib), 1042 patients did not receive NSAID

Lee et al. 200923

Duration of NSAID treatment

Surgical procedure

Type of study N (active/placebo)

References

No data

No fatalities in any group

Total number of deaths 19/11245 (0.2%)

1/750 (0.1%) vs. 0/251 (death from myocardial infarction)

No data

Death, NSAID vs. control

No data

Incidence of post-operative myocardial infarction was 1.3% for celecoxib, 0.8% for meloxicam/ketorolac, and 1.8% for patients who did not receive NSAIDs (NS)

10/11245 (0.09%), death from cardiac arrest

1/750 (0.1%) vs. 0/251 (death from myocardial infarction)

No data

Cardiovascular events NSAID vs. control

No data

No specific data

117/11245 (1%)

32/750 (4.3%) vs. 5/251 (2%), P = 0.07

No data

Surgical bleeding, NSAID vs. control

No data

No specific data

No data

15/750 vs. 3/251, NS

No data

Severe surgical bleeding, NSAID vs. control

No data

No specific data

4/11245 (0.03%)

1/750 (0.1%) vs. 0/251, NS

No data

GI-bleeding NSAID vs. control

No data

No specific data

10/11245 (0.09%)

0/750 vs. 0/251, NS

No cases of Post-operative renal failure requiring dialysis were described

Renal failure NSAID vs. control

No data

No specific data

No data

Creatinine clearance reduced by 16 ml/min (95% CI 5, 28) and potassium output by 38 mmol/day (95% CI 19, 56) with NSAID on the first day after surgery compared with placebo. No significant reduction in urine volume No data

Oliguria NSAID vs. control

Up to 115 h post-operatively

30-day incidence of anastomotic dehiscence as defined by authors

Peto OR = 2.16 (95% CI 0.85, 5.53), P = 0.11) in favour of the control group. Overall rates of anastomotic dehiscence were 14/272 (5.1%) in intervention arms vs. 5/208 (2.4%) in control arms. Comparable results for NSAIDs and COX-2 inhibitors

30 days post-operatively

2 weeks after last study medication

24–48 h after surgery

Follow-up period

No data

No data

No data

No data

Anastomotic leakage NSAID vs. control

Systematic reviews with meta-analyses, RCTs, observational cohort and retrospective studies, of perioperative NSAIDs/COX-2 inhibitors: adverse effects.

Table 3

Adverse effects of multimodal analgesia

5

6

Elective operation for colorectal cancer with colonic or rectal resection and primary anastomosis

Primary colorectal anastomosis

Neurosurgical, abdominal, urologic and orthopedic procedures, and endoscopies.

Cohort study based on prospective data; 622 patients received ibuprofen, 224 received diclofenac, 1871 patients did not receive NSAID

Retrospective study; 201 patients received non-selective NSAIDs, 79 patients received COX-2 inhibitors, 44 patients received both NSAID and COX-2 inhibitors, 471 patients received no NSAIDs

Meta-analysis of 27 RCTs comparing perior post-operative ketorolac to controls for bleeding. 1304 patients received ketorolac, 1010 patients in control group. Controls received: saline (16 studies), opioids (9 studies), acetaminophen (1 study), metamizole (1 study)

Klein et al. 201228

Gorissen et al. 201229

Gobble et al. 201430

Overall mortality rates were 13/324 (4.01%) in patients receiving NSAID and/or COX-2 inhibitor vs. 20/471 (4.2%) in patients receiving no NSAIDs (NS)

30-day post-operative mortality was 27/622 (4.1%) in the ibuprofen group, 4/224 (1.8%) in the diclofenac group, and 59/1871 (3.2%) in the control group (NS)

Death, NSAID vs. control

From 24 h to 4 No data days post-operatively. Doses: 7.5 to 60 mg. Time of administration: Intraoperatively in 18 studies, post-operatively in 9 studies.

Mean 2.8 (SD 1.3) days in the first 5 days after operation, starting 0.9 days after surgery

At least two days of NSAID treatment in the first seven days after surgery

Duration of NSAID treatment

No data

No data

No data

Cardiovascular events NSAID vs. control

7/27 trials reported of bleeding time. 2/7 studies: (NS) 5/7 trials reported prolonged bleeding time, but with no clinically significant difference in bleeding (NS)

No data

No data

Surgical bleeding, NSAID vs. control

Bleeding defined as: readmission, reoperation or requiring transfusion. Bleeding not increased with ketorolac (OR, 1.1; 95% CI: 0.61, 2.06; P = 0.72). Incidence: 33/1304 (2.5%), ketorolac group. 21/1010 (2.1%), control group. No difference between doses > 30 mg or < 30 mg

No data

No data

Severe surgical bleeding, NSAID vs. control

No data

No data

No data

GI-bleeding NSAID vs. control

No data

No data

No data

Renal failure NSAID vs. control

No data

No data

No data

Oliguria NSAID vs. control

No data

Leakage rates were 12.8% with diclofenac, 8.2% with ibuprofen, and 5.1% with no NSAID (P ≤ 0.004 for both NSAIDs vs. control). With multivariate logistic regression, diclofenac was a risk factor for leakage (OR = 7.2 (95% CI 3.8, 13.4), P < 0.001; Ibuprofen was not: OR 1.5 (0.8, 2.9), P = 0.18) Leakage rates were 13.2% with NSAID/ COX-2 inhibitors vs. 7.6% with no NSAID, OR = 1.84 (95% CI 1.13, 2.98), P = 0.01). This effect was mainly due to non-selective NSAID (14.5%, OR 2.13 (1.24, 3.66), P = 0.006); COX-2 inhibitors (9%, OR 1.16 (0.9, 2.75), P = 0.74)

Anastomotic leakage NSAID vs. control

CI, confidence interval; RCT, randomised clinical trial; NS, not significant; OR, odds ratio; Peto OR, peto odds ratio; NSAID, non-steroidal anti-inflammatory drug; COX-2, cyclooxygenase-2; SD, standard deviation.

Surgical procedure

Type of study N (active/placebo)

References

Table 3 Continued

Not stated

Not stated

Patients with available electronic medical records who had undergone an elective operation for colorectal cancer between 1 January 2006 and 31 December 2009. 30-day post-operative mortality

Follow-up period

O. Mathiesen et al.

Adverse effects of multimodal analgesia

fatalities may indicate that the patient populations in these studies are not representative for the general, surgical population, where substantial higher mortality rates have been reported.2,3,6 Anastomotic leakage. A meta-analysis of randomised clinical trials comparing a non-selective NSAID or COX-2 inhibitor with a control treatment in adults undergoing surgery with formation of at least one anastomosis of the small bowel, colon or rectum, showed a non-significantly increased risk {odds ratio (OR) [95% confidence interval (CI) 2.16 (0.9–5.5)]} of anastomotic leakage.27 This OR translated into a 5.1% vs. 2.4% leakage rate in the NSAID/COX-2 inhibitor and control group, respectively. Due to the small size of the included studies (480 patients in 6 studies), lack of heterogeneity in the results and the magnitude of the effect, the authors concluded that the non-significant results were due to a lack of power. Moreover, the included studies were designed to investigate the opioidsparing effect of NSAIDs with different COXselectivity. Therefore, incidence of anastomotic leakage was not a primary outcome in any of the studies.27 Other studies designed to investigate the risk of anastomotic leakage support an increased risk of leakage associated with the use of NSAIDs. In the cohort study of Klein et al.,28 evaluating 2756 patients having a colorectal anastomosis performed, diclofenac was a marked risk factor for leakage [OR (95% CI) 7.2 (3.8–13.4) P < 0.001] after adjusting for confounders. However, ibuprofen only showed a non-significant tendency [OR (95% CI): 1.5 (0.8–2.9), P = 0.18].28 The concern about diclofenac was raised as early as 2007, where this drug was introduced as part of the standard post-operative analgesic regimen in many surgical departments in Denmark. The anastomotic leakage rate for colonic resections was significantly increased from 3.9% to 20.5% after this introduction with no other possible explanation. The subsequent avoidance of diclofenac resulted in a decrease to 1.3%.31 Comparable results (2.3% vs. 21.2%) were found in patients having laparoscopic colorectal surgery without and with diclofenac.32 In a cohort study in 502 patients having colonic anastomoses performed,33 the leakage rate increased from 3.3% to 15.1% after the introduction of celecoxib as part of the standard analgesic regimen. After the discontinuation of celecoxib, the leakage rate dropped to 1.5%. In a retrospective study, leakage rates were 13.2% with NSAID/COX-2 inhibitors vs. 7.6% with no NSAID, [OR (95% CI) 1.8

(1.1–3.0), P = 0.01]. This effect was mainly due to non-selective NSAIDs: 14.5%, [OR (95% CI) 2.1 (1.2– 3.7), P = 0.006], whereas the effect observed with selective COX-2 inhibitors: 9%, [OR (95% CI) 1.2 (0.9–2.8), P = 0.74]29 was not statistically significant (Table 3). Bone healing. Concern has also been raised regarding delayed bone healing or even non-union of bone fracture following a NSAID-mediated COX-2 inhibition of prostacyclins. In a review from 2012, Kurmis et al.34 identified 316 potentially relevant studies on the effect of NSAIDs on bone fracture healing. Most studies were small, retrospective and/or cohort studies. In these studies, the vast amount of confounding variables was poorly controlled for, and the studies were most often inconclusive in relation to human NSAID use in the post-operative period. Although some animal studies indicate NSAID-mediated increased risk of bone fracture non-union, no prospective controlled human trials provide evidence to support this notion. It has been suggested that especially short NSAID treatments might be safe regarding bone healing.34

GCCs Data from systematic reviews and meta-analyses of analgesic efficacy. Two recent systematic reviews investigated the impact of systemic dexamethasone on post-operative analgesia and adverse effects.35,36 De Oliviera et al.35 included randomised clinical trials of a single perioperative i.v. dexamethasone administration (4–40 mg) with an inactive (placebo or ‘no treatment’) control group in mainly laparoscopic procedures, or small to moderate surgery. Reporting of adverse effects in the included studies was incomplete, but no statistical significant differences in wound infection or wound healing were observed (Table 4). Waldron et al.36 included trials of single-dose perioperative i.v. dexamethasone 1.25– 20 mg in adult patients undergoing mixed minor– major surgery. Again, data on AEs were sparse and incomplete, but no increase in wound infection nor delayed wound healing was observed with dexamethasone (Table 4). However, based on data from two of the included studies with 203 patients, blood sugar at 24 h post-operatively was marginally increased: 0.39 mmol/l [(95% CI) 0.04–0.74] with GCC.36 Note that these meta-analyses contain at least in part some of the same original trials that should be taken into consideration in the overall interpretation.

7

O. Mathiesen et al. Table 4 Systematic reviews with meta-analyses, of perioperative glucocorticoids in post-operative pain treatment: reported adverse effects. Ref

Drug and administration N (active/placebo)

Surgical population

Adverse effects

De Oliveira et al. 201135

Single-dose perioperative intravenous dexamethasone in adult patients. Comparisons were stratified by dose into three groups: low-dose (≤ 0.10 mg/kg), intermediate-dose (0.11–0.20 mg/kg) and high-dose (≥ 0.21 mg/kg) dexamethasone. A total of 2751 patients were analysed (24 trials)

Mixed minor–major surgery (not dental)

Waldron et al. 201236

Single-dose perioperative intravenous dexamethasone 1.25–20 mg in adult patients. 2997/2799 (45 trials)

Mixed minor–major surgery (not dental)

Wound infection: Moderate dose dexamethasone (0.11–0.20 mg/kg): RR = 0% (95% CI: −1.2, 1.2, NNH: lower CI = 83). NS Wound infection: High dose dexamethasone (> 0.2 mg/kg): RR = 0.3% (95% CI: −2.5, 3.1, NNH: lower CI = 32). NS Pruritus: OR = 0.72 (95% CI 0.2, 2.1). NS Wound infection: (14 trials, 1449 patients) RR = 0.63 (0.23, 1.69, P = 0.36). Wound healing: (9 trials, 1020 patients) RR = 1.27 (0.32, 4.96, P = 0.73). At 24 h post-operatively, blood sugar in the dexamethasone group was significantly higher: MD = 0.39 mmol/l (0.04, 0.74, P = 0.03) (2 trials, 203 patients)

RR, relative risk; NNH, numbers needed to harm; NS, not significant; CI, confidence interval; OR, odds ratio; MD, mean difference.

Data from other studies. A recent randomised clinical trial of 4494 patients undergoing cardiac surgery with cardiopulmonary bypass compared the effect of a single intraoperative dose of 1 mg/kg dexamethasone with placebo.37 The authors found no increase in major AEs, including death, myocardial infarction, stroke, respiratory failure, gastrointestinal bleeding or wound infection with dexamethasone within 30 days of randomisation.37 A marginal increase in blood sugar was reported with dexamethasone with the highest blood sugar during stay at the intensive care unit: 195 (50) vs. 177 (59) mg/dl in the dexamethasone vs. placebo groups, respectively (Table 5). In a meta-analysis of 54 randomised trials, with 3615 patients comparing GCC treatment to either placebo or no treatment in adult cardiac surgery,38 no increase in the incidence of AEs (mortality, myocardial- or pulmonary complications, renal failure and infections) was found with GCCs. However, incidence of gastrointestinal bleeding/ ulceration was increased in the GCC group (Table 5).38 Another meta-analysis of 51 randomised clinical trials compared high-dose methylprednisolone (15 mg/kg or 1 g administered as a single or repeated dose within a maximum of 3 days) with placebo or no treatment. The included trials were performed in elective cardiac surgery (28 trials), elective noncardiac surgery (13 trials), trauma surgery (6 trials) and spine surgery (4 trials). There was no

8

significant increase in complication rates. However, a non-significant increase in gastrointestinal bleeding [NNH = 333 (59; -)] and wound complications [NNH = 100 (38; -)] was observed (Table 5).39 Intraoperative dexamethasone 4–12 mg for postoperative nausea and vomiting (PONV) prophylaxis did not increase the rate or severity of post-operative wound complications in a retrospective study of 431 women undergoing laparotomy for endometrial cancer.40 Likewise, in a post hoc analysis of data from the PROXI trial (PeRioperative OXygen Fraction Effect on Surgical Site Infection and Pulmonary Complications After Abdominal Surgery) in 1386 patients who underwent emergency or elective laparotomy, perioperative administration of dexamethasone 4 mg was not significantly associated with surgical site infection or other wound-related complications (Table 5).41

Gabapentinoids The gabapentinoids, gabapentin and pregabalin, share a principal mechanism of action, blocking the alpha-2-delta subunits of voltage-sensitive calcium channels in pre-synaptic afferent neurons. However, differences in pharmacokinetics and possibly different profile of other mechanisms of action such as e.g., effects on N-Methyl-D-aspartate receptors, alpha-2-adrenergic receptors or adenosine A1 receptors, may explain the differences in the adverse effect profiles of these two drugs.

Cardiac (28 trials) and non-cardiac (13 trials) surgery, Trauma surgery (6 trials), spine surgery 4 trials).

Systematic Sauerland et al. 200039 review with meta-analyses of 51 trials 15 mg/kg or 1 g methyl-prednisolone given as a single or repeated dose within a maximum of 3 days

Single or multiple high doses of different GCCs: dexamethasone, methyl-prednisolone, solumedrol, hydrocorticone and prednisolone.

Cardiac surgery

Dose, duration of GCC-treatment and follow-up period

Systematic Dieleman et al. 201138 review with meta-analyse of 54 randomised clinical trials with 3615 patients

Surgical procedure

Dexamethasone 1 mg/kg (max 100 mg) single dose, 1 year follow-up, but only data for 30days follow-up period in trial.

Type of study N (active /placebo)

Randomised Cardiac Dieleman surgery et al. 201237 controlled multicentre study 2235 patients received dexamethasone, 2247 received placebo

Ref Cardio-vascular complications GCC vs. control

26 trials with 2103 patients: Peto OR = 0.95 (95% CI 0.57, 1.60). NS Atrial fibrillation 17 studies with 1399 patients Peto OR 0.60 (95% CI 0.46, 0.78) (reduced with GCC)

50 trials. No data 20/1170 vs. 32/1190. RD: −0.9% (95% CI: −2.5, 0.6). NNT = 111. NS

49 trials with 3213 patients: Peto OR = 1.06 (95% CI 0.58, 1.95) NS

31/2235 Myocardial (1.4%) vs. infarction 34/2247 35/2235 (1.6%) (1.5%) RR vs. 39/2247 = 0.92 (95% CI (1.7%) RR = 0.90 0.57, 1.49) (95% CI 0.57, 1.42) Stroke 29/2235 (1.3%) vs. 32/2247 (1.4%) RR = 0.91 (95% CI 0.55, 1.50) Atrial fibrillation 739/2235 (33.1%) vs. 790/2247 35.2%) RR = 0.94 (95% CI 0.87, 1.02, P = 0.14)

Death, GCC vs. control

35 trials. 74/981 vs. 115/1011. RD –3.5% (95% CI −6.1, −1.0). NNT = 29 (95% CI 16, 100)

21 trials with 1340 patients: Peto OR 0.83 (95% CI 0.49, 1.40) NS

67/2235 (3.0%), vs. 97/2247 (4.3%) RR = 0.69 (95% CI 0.51, 0.94, P = 0.02)

Pulmonary complications, GCC vs. control

No data

13 trials with 825 patients: Peto OR 1.0 (95% CI 0.45, 2.19). NS

No data

Blood transfusion (yes/no) 6 trials with 535 patients: Peto OR 0.87 (95% CI 0.54, 1.39). NS

No data

Wound healing, GCC vs. control

Gastrointestinal bleeding. 34 trials. 10/973 vs. 7/1003 RD +0.3% (95% CI −1.0, 1.7) NNH = 333 (95% CI 59, -). NS

No data

4 trials with No data 304 patients: Peto OR 2.84 (95% CI 0.40, 20.36) for bleeding/ulceration. (increased with GCC)

Gastrointestinal bleeding 13/2235 (0.6%) vs. 11/2247 (0.5%) RR = 1.19 (95% CI 0.53, 2.65, P = 0.67)

Bleeding GCC Gastrointestinal vs. control complications GCC vs. control

According to No data RIFLE criteria. 28/2235 (1.3%) vs. 40/2247 (1.8%) RR = 0.70 (95% CI 0.44, 1.14)

Renal failure GCC vs. control

No analysis possible

Highest blood sugar at the ICU (Mean and SD, mg/ml) 195 (50) vs. 177 (59) (P < 0.001)

Blood sugar, GCC vs. control

Wound No data complication (infection, bleeding, leakage, seroma) 34 trials: 28/972 vs. 20/997 patients. RD: +1.0 (95% CI −7, 2.6), NNH = 100 (95% CI 38, -). NS

Infectious complications 16 trials with 1517 patients: Peto OR 0.86 (95% CI 0.56, 1.31). NS

Wound infection 34/2235 (1.5%) vs. 32/2247 (1.4%) RR = 1.07 (95% CI 0.66, 1.72, P = 0.79) Any post-operative infection 212/2235 (9.5%) vs. 333/2247 (14.8%) RR = 0.64 (95% CI 0.54, 0.75, P < 0.001)

Wound infection, GCC vs. control

Systematic reviews with meta-analyses, RCTs, observational cohort and retrospective studies of perioperative glucocorticoids: reported adverse effects.

Table 5

Length of hospital stay (days), 15 trials with 635 patients: Peto OR −0.59 (95% CI −0.84, −0.34) NS ICU length of stay 25 trials with 1215 patients: Peto OR–2.32 h (95% CI −2.84, −1.81) (improved with GCC) Time to extubation, re-thoracotomy and neurological complications: NS 4 trials: Length of stay: −1.9 days (95% CI −6.1, 2.4)

Delirium 205/2235 (9.2%) vs. 262/2247 (11.7%) RR = 0.79 (95% CI 0.66, 0.94, P = 0.006) Length of hospital stay 8 (7 to 13) days vs. 9 (7 to 13) days (P < 0.009)

Other outcomes, GCC vs. control

Adverse effects of multimodal analgesia

9

10

Open hysterectomy for endometrial cancer

Emergency or elective laparatomy

Register study 192 patients received PONV prophylaxis and 239 patients did not.

Post hoc analysis of randomised study (PROXItrial)

Bolac et al. 201340

Dahl et al. 201441

Dexamethasone

Dexamethasone 4–12 mg single-dose perioperatively as PONV prophylaxis. Patients followed until healing, lost to follow-up or death

Dose, duration of GCC-treatment and follow-up period

30-day mortality: 10/412 (2.4%) vs. 37/895 (4.1%). OR 1.06 (95% CI 0.47, 2,40, P = 0.88) Long term mortality: 114/412 (28%) vs. 257/895 (29%). OR 0.95 (95% CI 0.7, 1.3, P = 0.76)

No data

Death, GCC vs. control

No data

No data

Cardio-vascular complications GCC vs. control No data

Renal failure GCC vs. control

Pneumonia No data 15/412 (3.6%) vs. 65/895 (7.3%) OR 0.7 (95% CI 0.36, 1.34, P = 0.28) Respiratory failure 14/412 (3.4%) vs. 54/895 (6%) OR 0.81 (95% CI 0.41, 1.61, P = 0.55)

No data

Pulmonary complications, GCC vs. control

No data

No data

Anastomotic leak 6/412 (1.5%) vs. 23/895 (2.6%). OR 0.91 (95% CI 0.34, 2.43, P = 0.84)

No data

Burst abdomen 9/412 (2.2%) vs. 28/895 (3.1%) OR 0.61 (95% CI 0.25, 1.49, P = 0.28) Wound dehiscence 12/412 (2.9%) vs. 35/895 (3.9%) OR 0.7 (95% CI: 0.33, 1.52, P = 0.37)

No difference in time to wound healing in patients with infection. (P = 0.48)

Bleeding GCC Gastrointestinal Wound vs. control complications GCC healing, GCC vs. control vs. control

Blood sugar, GCC vs. control

Composit No data endpoint: surgical site infection (SSI), anastomotic leak, wound dehiscence, burst abdomen and 30-day mortality. 88/412 (21%) vs. 249/895 (28%). OR 0.9 (95% CI 0.65, 1.24, P = 0.52)

Included: No data superficial surgical site infection, cellulitis, wound separation and fascial dehiscence. 53/192 (27.6%) vs. 81/239 (33.9%) OR: 0.74 (95% CI: 0.49, 1.13, P = 0.16). Incidence not associated to dose

Wound infection, GCC vs. control

Sepsis 5/412 (1.2%) vs. 29/895 (3.2%) OR: 0.56 (95% CI 0.2, 1.61, P = 0.28)

Other outcomes, GCC vs. control

GCC, glucocorticoids; RR, relative risk; CI, confidence interval; ICU, intensive care unit; Peto OR, peto odds ratio; OR, odds ratio; NS, not significant; RD, risk difference; NNT, numbers needed to treat; NNH, numbers needed to harm; PONV, post-operative nausea and vomiting; RIFLE, Risk, Injury, and Failure; and Loss; and End-stage kidney disease.

Surgical procedure

Type of study N (active /placebo)

Ref

Table 5 Continued

O. Mathiesen et al.

Adverse effects of multimodal analgesia

Data from systematic reviews and meta-analyses of analgesic efficacy of gabapentin. A number of systematic reviews with meta-analyses address the post-operative analgesic effect of gabapentin.42–49 Most of these reviews have been published 6–8 years ago and include 4–21 RCTs, which currently only represent a fraction of the RCTs published. Furthermore, most of the published meta-analyses include at least partly the same original trials, which should be taken into consideration in the overall interpretation. All reviews focus on gabapentin-mediated reductions of opioid-related adverse effects. The reviews only have data on a limited number of gabapentinrelated adverse effects: sedation, dizziness, lightheadedness, visual disturbances and headache. The most common adverse effects reported with a single-dose gabapentin treatment are sedation [(NNH = 8)47 and (NNH = 35)45], and dizziness (NNH = 12).47 In three of eight reviews, gabapentin adverse effects were reported being not different from placebo (Table 6). Data from other studies on gabapentin. The above findings of adverse effects of short-term gabapentin treatment are in concordance with those presented in a 2011 Cochrane review of gabapentin treatment (≥ 1200 mg/day) of chronic neuropathic pain.50 Based on data from 11 trials with 2356 patients, the overall pooled estimate of an effect on experiencing one or more adverse effects with gabapentin as compared with placebo, was a risk ratio (RR) = 1.3 (95% CI 1.2–1.4, NNH 6.6). The most frequently reported AEs from longer-term gabapentin treatment included sedation [RR = 3.2 (95% CI 2.5–4.2, NNH 9.2)] and dizziness [RR = 3.2 (95% CI 2.5–4.2, NNH 7.0)] followed by peripheral oedema (RR = 3.4 (95% CI 2.1–5.3, NNH 19) and ataxia (RR = 4.5 (95% CI 1.9–11, NNH 13). Eleven per cent of the patients withdrew from gabapentin treatment because of AEs.50 Data from systematic reviews and meta-analyses of analgesic efficacy of pregabalin. Two recent metaanalyses of pregabalin for post-operative pain treatment included 1151 and 1852 randomised trials, respectively. Reported adverse effects of pregabalin treatment were visual disturbances51,52 (NNH = 6),51 as well as dizziness and light-headedness.52 Interestingly, the relative risk of visual disturbances was approximately the same for both low and high doses of pregabalin (Table 6). The others who reported

adverse effects, headache and lack of concentration were not different from placebo. Data from other studies on pregabalin. A recent metaanalyses based on 38 randomised trials with 11.918 individuals, presented the adverse effect profile of at least 4 weeks of pregabalin treatment for different chronic neurologic and psychiatric disorders.53 A total of 20 AEs were significantly associated with pregabalin treatment, including dizziness, ataxia, blurred vision, tremor, somnolence, confusion, abnormal thinking, euphoria, asthenia, fatigue, oedema, dry mouth and constipation. Sixteen of these were related to affected cognition and/or motor coordination.53 These AEs may therefore also be relevant to consider with prolonged postoperative pregabalin usage.

Adverse effects of drug combinations Combinations of NSAIDs and paracetamol Data from systematic reviews and meta-analyses of analgesic efficacy. Four studies54–57 investigated the analgesic effect of concomitant administration of paracetamol and NSAID vs. either drug alone in post-operative pain treatment. Reporting of adverse effects was sparse with no systemically reported data in two studies,54,56 and in a third study,55 no adverse effects were ascribed to the combination treatment in 14 of 15 included trials; one trial supplied no adverse effect information. In the review by Derry et al.,57 paracetamol and ibuprofen treatment was investigated in three trials with 1647 patients undergoing third molar surgery. The reported adverse effects were related to the type of surgery: swelling of the face, nausea, vomiting, headache and dizziness. Based on data from two of three trials in this review, the incidence of at ‘least one adverse effect’ was reduced with the combination of ibuprofen and paracetamol [37% (80/216, range: 23% to 69%] compared with ibuprofen treatment alone [55% (78/143, range: 35% to 75%]. Data from other studies. In a longitudinal cohort study, with data of patients from the UK General Practice Database, safety outcomes of the combination treatment of ibuprofen and paracetamol vs. either drug alone were investigated.58 Outcome measures were gastrointestinal events, myocardial infarction, stroke, renal failure, congestive heart failure, overdose, suicidal behaviour and mortality. The population included nearly 1.2 million adult

11

O. Mathiesen et al. Table 6 Systematic reviews with meta-analyses, of perioperative gabapentinoids in post-operative pain treatment: Reported adverse effects. Ref Gabapentin Straube et al. 201042 Seib and Paul 200643 Hurley et al. 200644

Drug and administration N (active/placebo)

Surgical population

Adverse effects

Single-dose gabapentin 250–500 mg 198/172 (4 trials)

4 unpublished studies: 3 in dental surgery, one in major orthopaedic surgery Mixed moderate–major surgery

Type of AE no specified, but not different from placebo. In three studies of gabapentin 250 mg vs. placebo: RR = 0.91 (95% CI: 0.66, 1.25). No serious adverse event with gabapentin. Dizziness, light-headedness, visual disturbance, headache: not different from placebo

Mixed minor–major surgery

Incidence of sedation increased: OR = 3.28; (95% CI: 1.21, 8.87); Dizziness/light-headedness: not different from placebo: OR = 1.27; (95% CI: 0.67, 2.41) Incidence of sedation increased: OR = 3.86; (95% CI: 2.50, 5.94, P < 0.00001); NNH: 8. Dizziness not different: OR 1.34 (95% CI 0.86, 2.1, P = 0.2) Dizziness significantly increased (RR = 1.4, 95% CI: 1.06, 1.84, P = 0.02). Sedation borderline significant: (RR = 1.65; 95% CI: 1.0, 2.4, P = 0.05) Sedation NNH = 35; dizziness NNH = 12 (for single-dose gabapentin)

Single- and multiple-dose gabapentin 300–1200 mg 333/330 (8 trials) Single- and multiple-dose gabapentin 300–1200 mg 449/447 (12 trials)

Ho et al. 200645

Single- and multiple-dose gabapentin 300–1200 mg 614/537 (16 trials)

Mixed minor–major surgery

Peng et al. 200746

Single- and multiple-dose gabapentin 300–1200 mg A total of 1181 patients were included (18 trials) Single- and multiple-dose gabapentin 300–1200 mg 786/? (21 trials) Single- and multiple-dose gabapentin 300–1200 mg Hysterectomy 137/134 (5 trials); Spinal surgery 113/73 (4 trials)

Mixed minor–major surgery

Single and multiple dose of gabapentinoids (gabapentin 300–1200 mg; and pregabalin 150–300 mg); 218/216 (7 trials)

Spinal surgery

Single- and multiple-dose pregabalin 50–600 mg 521/378 (11 trials) Single- and multiple-dose pregabalin 50–600 mg 850/697 (18 trials )

Mixed minor–major surgery

Tiippana et al. 200747 Mathiesen et al. 200748

Yu et al. 201349

Pregabalin Zhang et al. 201151 Engelman and Cateloy 201152

Mixed minor–major surgery Open hysterectomy/spinal surgery

Mixed minor–major surgery

Dizziness: NS. [Open hysterectomy: RR = 1.4, (95% CI: 0.8, 7.2, P = 0.15)]; Spinal surgery: RR = 1,31, (95% CI: 0.6, 3.1, P = 0.53)] Sedation: NS. [Open hysterectomy: RR = 2.3, (95% CI: 0.8, 7.2, P = 0.14); Spinal surgery: RR = 1.5, (95% CI: 0.3, 8.6, P = 0.65)] ‘Gabapentin and pregabalin could be well tolerated in the included trials’

Visual disturbance increased: RR 3.29; (95% CI: 1.95, 5.57, NNH = 6) Sedation, dizziness, headache: NS Sedation: RR = 1.46, (95% CI 0.96, 2.23, P = 0.08). Dizziness/light-headedness increased: RR = 1.43, (95% CI: 1.17, 1.74, P = 0.0004) Visual disturbances increased [Doses 50–150 mg: RR = 3.87 (95% CI: 1.05, 14.17, P = 0.04). Doses 225–300 mg: RR = 3.51, (95% CI: 1.16, 10.61, P = 0.03) Doses 600–750 mg: RR = 4.02, (95% CI 1.43, 11.33, P = 0.008)] Lack of concentration/confusion: NS

RR, relative risk; CI, confidence interval; OR, odds ratio; NNH, numbers needed to harm; NS, not significant.

patients who received either paracetamol or ibuprofen, and more than 13,000 patients who received the combination. No increased risk for the safety outcomes was found for the combination treatment compared with either treatment alone in this study.58

data on adverse effect of these combinations. Consequently, no systematic knowledge of potential harms/adverse effects of the combination of three or four of the most commonly used analgesics in post-operative pain management are available.

Combinations of paracetamol, NSAIDs, GCCs and gabapentinoids

Discussion

We found no systematic reviews, meta-analyses, randomised trials or cohort studies that presented

The main findings of the present review is that most data on adverse effects of the commonly used anal-

12

Adverse effects of multimodal analgesia

gesics originate from trials with mainly small study populations, and studies where adverse effects were not primary outcome measures. Furthermore, in most of the studies, observations were limited to the immediate post-operative period, typically 24 h. Even if trials report low incidences of adverse effects, inclusion and exclusion criteria of clinical analgesic trials are often very restrictive, and consequently, the incidence of adverse effects does not necessarily match that of general surgical populations. This may be particularly problematic in NSAID/COX-2 studies, due to the recent information on the risks of NSAIDs on the cardiovascular system and putative effect on bone healing or anastomotic leakage in large bowel surgery. Based on the present data, it is therefore of concern that a potential additive pain-relieving effect of multimodal analgesia may inherently include also a risk of an increased incidence of (serious) AEs. Paracetamol is very commonly used in postoperative pain management and several reviews have demonstrated the potential of paracetamol to reduce both opioid consumption and PONV. These reviews are based on studies were information on adverse effects was either sparse or described as mild and transient and not different from placebo11–16 (Table 1). However, the majority of the trials of oral paracetamol were published more than 2 decades ago, and the description of the follow-up period was unclear and typically ranging from 4 h to 7 days. Despite its frequent use, the mode of action of paracetamol is still unclear. It is believed to act centrally via cannabinoid, vanilloid or descending serotonergic pain-modulating systems,59 but some recent data advocate for a COX inhibitory effect on both central and peripheral mediated prostaglandin synthesis.59 Cardiovascular and gastrointestinal adverse effects as well as platelet dysfunction may therefore theoretically be of concern, especially in combination with definite COX inhibitors.17 In a recent and large epidemiological study of patients from general practice, however, concomitant use of paracetamol and ibuprofen did not increase adverse effects compared with the use of either drug alone.58 Future randomised trials of post-operative pain should therefore focus on possible prolonged adverse effects from the use of paracetamol alone and in combination with NSAIDs.17 NSAIDs and the selective COX-2 inhibitors are also very commonly used analgesics in combination regimens, but they are also some of the most controversial drugs in post-operative pain treatment. The pain-relieving effects of NSAIDs/COX-2 inhibi-

tors are well documented and include a capacity to reduce opioid-related adverse effects associated with their opioid-sparing ability.14,15 Whereas selective COX-2 inhibitors inhibit the production of prostacyclin without affecting the production of thromboxane A2, which is COX-1 mediated, traditional NSAIDs have both COX-1 and COX-2 affinity and thereby inhibit both the production of prostacyclin and thromboxane A2. Prostacyclin inhibits platelet aggregation, prevents the proliferation of vascular smooth-muscle cells and causes vasodilatation. Thromboxane A2 is the main COX-1-mediated product of platelets and causes platelet aggregation, vasoconstriction and vascular proliferation.60 Thus, selective COX-2 inhibitors may be thrombogenic60 whereas aspirin-like protective effects have been attributed to the traditional NSAIDs such as naproxen and ibuprofen in some studies.61 Potential adverse effects, such as fatal cardiovascular events,62 gastrointestinal bleeding,63 colorectal anastomotic leakage,27–29,64 surgical bleeding,14 renal impairment23 and impaired fracture healing,34 may be associated with non-selective NSAIDs and the selective COX-2 inhibitors. On the other hand, traditional NSAIDs may also have beneficial effects in the surgical patient, due to their aspirin-like effects,61 although a recent trial of perioperative aspirin (The Perioperative Ischemic Evaluation-2 trial65) did not demonstrate a significant reduction of the composite outcome measure of death or myocardial infarction, but instead an increased risk of major bleeding. The consequences of a perioperative, short-lasting (e.g., 5 days) regimen of traditional NSAIDs on longterm (3–6 months) morbidity and mortality in patients undergoing major surgery have not been assessed in large randomised clinical trials. Furthermore, the possible risk associated with combinations with other non-opioid analgesics needs evaluation in large trials. The anti-inflammatory effects of GCCs are probably the main reason for their analgesic effects,66 although an early antihyperalgesic effect also has been demonstrated.67 Furthermore, GCCs reduce post-operative nausea, vomiting and possibly fatigue and dizziness in the recovery period.68,69 Thus, GCCs may have the potential to be an important part of post-operative analgesic treatment. However, potential risks of GCC treatment including wound infection, delayed wound healing and hyperglycaemia are of concern. These adverse effects are known to be related to chronic GCC treatment. No differences in incidence of wound infection and delayed wound

13

O. Mathiesen et al.

healing from a single dose of GCC was discovered in the studies included in the present review. These studies include two systematic reviews of singledose perioperative GCCs,35,36 a register study of 431 gynaecological surgery patients,40 a secondary analysis of the PROXI trial in 1386 patients undergoing laparotomy41 and a multicentre study of high-dose dexamethasone in 4494 cardiac surgery patients.37 It must be considered, although, that adverse effects are often under-reported and that trials frequently suffer from insufficient methodology to detect these problems. The hyperglycaemic response to GCCs is modest, at most.37 So far, there are no solid evidence for the association between a single dose of GCCs and raised blood sugar and deteriorated postoperative outcome,70 although a recent editorial highlighted the need for collecting more data from proper studies on this issue.71 Several meta-analyses indicate that the gabapentinoids are potent non-opioid analgesics in acute post-operative pain.47,48 Gabapentin and pregabalin reduces pain and opioid requirements and attenuate opioid-related adverse effects. The present review, however, points to an increased incidence of adverse effects, in particular sedation, dizziness and visual disturbances, although the clinical relevance of these effects is not clear. It should be noted that post-operative dizziness is a well-know complication, especially for the elderly, which potentially could be worsened with gabapentinoids. Furthermore, the wide spectrum of potential adverse effects of gabapentinoids, e.g., sedation, dizziness, headache, oedema, visual and cognitive disturbances and ataxia, which are well known from chronic pain studies,50,53 is not sufficiently investigated in relation to acute pain treatment. Therefore, further large-scale trials are needed with focus on both beneficial and harmful effects. Likewise, future RCTs must address a broader range of adverse effects, and especially, their clinical relevance. Hopefully, large dose-response trials will be able to address the unanswered question of the optimal balance between efficacy and adverse effects. Finally, although gabapentin and pregabalin have the same presumed mode of action, the literature points toward a different adverse effect profiles with e.g. visual disturbances being more pronounced with pregabalin.51,52 Consequently, comparative studies of gabapentin and pregabalin investigating analgesic efficacy and relevant adverse effects are needed. Systematic reviews, meta-analyses or cohort trials with reliable information of (serious) adverse effects

14

from perioperative use of different combinations of paracetamol, NSAIDs, GCCs and gabapentinoids were sparse or not retrievable. Little information is available about the combined effect of paracetamol and NSAIDs on prostacyclin inhibition and the potential adverse effects such as cardiovascular events, bone healing and anastomotic leakage. Furthermore, systematic knowledge of the effect of a further addition of an anti-inflammatory drug like dexamethasone to an NSAID or to paracetamol and NSAID is not available. It is therefore possible that multimodal analgesia may expose patients to an increased risk of AEs. As multimodal therapy is the leading principle within post-operative pain treatment, it is of substantial concern that our knowledge of potential escalation of adverse effects from the most common analgesic combinations is deficient. For post-operative patients, after millions of surgical interventions worldwide, safety of the medical intervention is a crucial focus, and following the Vioxx scandal, it is important to emphasise that even few weeks intervention with e.g., selective COX-2 inhibitors, may cause serious harm to patients.72,73 Therefore, future trials need to focus not only on analgesic effects, but also on prolonged follow-up of possible (serious) adverse effects from perioperative combinations of analgesics. Especially, trials investing safety issues of short-term perioperative NSAID treatment with ‘pre-rofecoxib’ drugs, as ibuprofen and naproxen, are needed to ensure that patient safety is not compromised by the choice of postoperative analgesic treatment. In conclusion, effective perioperative pain management is crucial for the surgical patients’ wellbeing and optimal rehabilitation after surgery. Combinations of non-opioid analgesics may improve pain relief due to additive or even synergistic effects. Despite this potential, reporting of AEs associated with the currently applied non-opioid analgesics, including their combinations, is generally sparse and insufficient, and assessment is most often confined to the immediate perioperative period. Therefore, our knowledge of benefit and harm related to multimodal pain treatment is deficient and needs clarification in large trials with prolonged observation. Conflicts of interest: None. Funding: None.

References 1. Weiser TG, Regenbogen SE, Thompson KD, Haynes AB, Lipsitz SR, Berry WR, Gawande AA. An estimation of the

Adverse effects of multimodal analgesia

2.

3.

4.

5. 6.

7. 8. 9. 10.

11. 12.

13.

14.

15.

16.

global volume of surgery: a modelling strategy based on available data. Lancet 2008; 372: 139–44. Noordzij PG, Poldermans D, Schouten O, Bax JJ, Schreiner FA, Boersma E. Postoperative mortality in the Netherlands: a population-based analysis of surgery-specific risk in adults. Anesthesiology 2010; 112: 1105–15. Devereaux PJ, Chan MT, Alonso-Coello P, Walsh M, Berwanger O, Villar JC, Wang CY, Garutti RI, Jacka MJ, Sigamani A, Srinathan S, Biccard BM, Chow CK, Abraham V, Tiboni M, Pettit S, Szczeklik W, Lurati Buse G, Botto F, Guyatt G, Heels-Ansdell D, Sessler DI, Thorlund K, Garg AX, Mrkobrada M, Thomas S, Rodseth RN, Pearse RM, Thabane L, McQueen MJ, VanHelder T, Bhandari M, Bosch J, Kurz A, Polanczyk C, Malaga G, Nagele P, Le Manach Y, Leuwer M, Yusuf S. Association between postoperative troponin levels and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2012; 307: 2295–304. Devereaux PJ, Goldman L, Cook DJ, Gilbert K, Leslie K, Guyatt GH. Perioperative cardiac events in patients undergoing noncardiac surgery: a review of the magnitude of the problem, the pathophysiology of the events and methods to estimate and communicate risk. CMAJ 2005; 173: 627–34. Aitkenhead AR. Injuries associated with anaesthesia. A global perspective. Br J Anaesth 2005; 95: 95–109. Pearse RM, Moreno RP, Bauer P, Pelosi P, Metnitz P, Spies C, Vallet B, Vincent JL, Hoeft A, Rhodes A. Mortality after surgery in Europe: a 7 day cohort study. Lancet 2012; 380: 1059–65. Kehlet H, Dahl JB. Anaesthesia, surgery, and challenges in postoperative recovery. Lancet 2003; 362: 1921–8. Kehlet H, Wilmore DW. Evidence-based surgical care and the evolution of fast-track surgery. Ann Surg 2008; 248: 189–98. Dahl JB, Mathiesen O, Kehlet H. An expert opinion on postoperative pain management, with special reference to new developments. Expert Opin Pharmacother 2010; 11: 2459–70. Dahl JB, Nielsen RV, Wetterslev J, Nikolajsen L, Hamunen K, Kontinen VK, Hansen MS, Kjer JJ, Mathiesen O; Scandinavian Postoperative Pain Alliance (ScaPAlli). Post-operative analgesic effects of paracetamol, NSAIDs, glucocorticoids, gabapentinoids and their combinations: a topical review. Acta Anaesthesiol Scand 2014. doi: 10.1111/aas.12382. Toms L, McQuay HJ, Derry S, Moore RA. Single dose oral paracetamol (acetaminophen) for postoperative pain in adults. Cochrane Database Syst Rev 2008; (4): CD004602. Tzortzopoulou A, McNicol ED, Cepeda MS, Francia MB, Farhat T, Schumann R. Single dose intravenous propacetamol or intravenous paracetamol for postoperative pain. Cochrane Database Syst Rev 2011; (10): CD007126. Remy C, Marret E, Bonnet F. Effects of acetaminophen on morphine side-effects and consumption after major surgery: meta-analysis of randomized controlled trials. Br J Anaesth 2005; 94: 505–13. Elia N, Lysakowski C, Tramer MR. Does multimodal analgesia with acetaminophen, nonsteroidal antiinflammatory drugs, or selective cyclooxygenase-2 inhibitors and patientcontrolled analgesia morphine offer advantages over morphine alone? Meta-analyses of randomized trials. Anesthesiology 2005; 103: 1296–304. McDaid C, Maund E, Rice S, Wright K, Jenkins B, Woolacott N. Paracetamol and selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDS) for the reduction of morphine-related side effects after major surgery: a systematic review. Health Technol Assess 2010; 14: 1–153. Apfel CC, Turan A, Souza K, Pergolizzi J, Hornuss C. Intravenous acetaminophen reduces postoperative nausea and vomiting: a systematic review and meta-analysis. Pain 2013; 154: 677–89.

17. Hinz B, Brune K. Paracetamol and cyclooxygenase inhibition: is there a cause for concern? Ann Rheum Dis 2012; 71: 20–5. 18. Doherty M, Hawkey C, Goulder M, Gibb I, Hill N, Aspley S, Reader S. A randomised controlled trial of ibuprofen, paracetamol or a combination tablet of ibuprofen/ paracetamol in community-derived people with knee pain. Ann Rheum Dis 2011; 70: 1534–41. 19. Moore RA, Derry S, McQuay HJ, Wiffen PJ. Single dose oral analgesics for acute postoperative pain in adults. Cochrane Database Syst Rev 2011; (9): CD008659. 20. Bainbridge D, Cheng DC, Martin JE, Novick R. NSAIDanalgesia, pain control and morbidity in cardiothoracic surgery. CMAJ 2006; 53: 46–59. 21. De Oliveira GS Jr., Agarwal D, Benzon HT. Perioperative single dose ketorolac to prevent postoperative pain: a metaanalysis of randomized trials. Anesth Analg 2012; 114: 424– 33. 22. Jirarattanaphochai K, Jung S. Nonsteroidal antiinflammatory drugs for postoperative pain management after lumbar spine surgery: a meta-analysis of randomized controlled trials. J Neurosurg Spine 2008; 9: 22–31. 23. Lee A, Cooper MG, Craig JC, Knight JF, Keneally JP. Effects of nonsteroidal anti-inflammatory drugs on postoperative renal function in adults with normal renal function. Cochrane Database Syst Rev 2007; (2): CD002765. 24. Merry AF, Webster CS, Holland RL, Middleton NG, Schug SA, James M, McGrath KA. Clinical tolerability of perioperative tenoxicam in 1001 patients–a prospective, controlled, double-blind, multi-centre study. Pain 2004; 111: 313–22. 25. Forrest JB, Camu F, Greer IA, Kehlet H, Abdalla M, Bonnet F, Ebrahim S, Escolar G, Jage J, Pocock S, Velo G, Langman MJ, Bianchi PG, Samama MM, Heitlinger E. Ketorolac, diclofenac, and ketoprofen are equally safe for pain relief after major surgery. Br J Anaesth 2002; 88: 227–33. 26. Liu SS, Bae JJ, Bieltz M, Ma Y, Memtsoudis S. Association of perioperative use of nonsteroidal anti-inflammatory drugs with postoperative myocardial infarction after total joint replacement. RAPM 2012; 37: 45–50. 27. Burton TP, Mittal A, Soop M. Nonsteroidal antiinflammatory drugs and anastomotic dehiscence in bowel surgery: systematic review and meta-analysis of randomized, controlled trials. Dis Colon Rectum 2013; 56: 126–34. 28. Klein M, Gogenur I, Rosenberg J. Postoperative use of non-steroidal anti-inflammatory drugs in patients with anastomotic leakage requiring reoperation after colorectal resection: cohort study based on prospective data. BMJ 2012; 345: e6166. 29. Gorissen KJ, Benning D, Berghmans T, Snoeijs MG, Sosef MN, Hulsewe KW, Luyer MD. Risk of anastomotic leakage with non-steroidal anti-inflammatory drugs in colorectal surgery. Br J Surg 2012; 99: 721–7. 30. Gobble RM, Hoang HL, Kachniarz B, Orgill DP. Ketorolac does not increase perioperative bleeding: a meta-analysis of randomized controlled trials. Plast Reconstr Surg 2014; 133: 741–55. 31. Rosenberg J, Harvald T. Severe complications with diclofenac after colonic resection. Dis Colon Rectum 2007; 50: 685. 32. Klein M, Andersen LP, Harvald T, Rosenberg J, Gogenur I. Increased risk of anastomotic leakage with diclofenac treatment after laparoscopic colorectal surgery. Dig Surg 2009; 26: 27–30. 33. Holte K, Andersen J, Jakobsen DH, Kehlet H. Cyclooxygenase 2 inhibitors and the risk of anastomotic leakage after fast-track colonic surgery. Br J Surg 2009; 96: 650–4.

15

O. Mathiesen et al. 34. Kurmis AP, Kurmis TP, O’Brien JX, Dalen T. The effect of nonsteroidal anti-inflammatory drug administration on acute phase fracture-healing: a review. J Bone Joint Surg Am 2012; 94: 815–23. 35. De Oliveira GS Jr., Almeida MD, Benzon HT, McCarthy RJ. Perioperative single dose systemic dexamethasone for postoperative pain: a meta-analysis of randomized controlled trials. Anesthesiology 2011; 115: 575–88. 36. Waldron NH, Jones CA, Gan TJ, Allen TK, Habib AS. Impact of perioperative dexamethasone on postoperative analgesia and side-effects: systematic review and meta-analysis. Br J Anaesth 2013; 110: 191–200. 37. Dieleman JM, Nierich AP, Rosseel PM, van der Maaten JM, Hofland J, Diephuis JC, Schepp RM, Boer C, Moons KG, van Herwerden LA, Tijssen JG, Numan SC, Kalkman CJ, van Dijk D. Intraoperative high-dose dexamethasone for cardiac surgery: a randomized controlled trial. JAMA 2012; 308: 1761–7. 38. Dieleman JM, van Paassen J, van Dijk D, Arbous MS, Kalkman CJ, Vandenbroucke JP, van der Heijden GJ, Dekkers OM. Prophylactic corticosteroids for cardiopulmonary bypass in adults. Cochrane Database Syst Rev 2011; (5): CD005566. 39. Sauerland S, Nagelschmidt M, Mallmann P, Neugebauer EA. Risks and benefits of preoperative high dose methylprednisolone in surgical patients: a systematic review. Drug Saf 2000; 23: 449–61. 40. Bolac CS, Wallace AH, Broadwater G, Havrilesky LJ, Habib AS. The impact of postoperative nausea and vomiting prophylaxis with dexamethasone on postoperative wound complications in patients undergoing laparotomy for endometrial cancer. Anesth Analg 2013; 116: 1041–7. 41. Dahl RM, Wetterslev J, Jorgensen LN, Rasmussen LS, Moller AM, Meyhoff CS. The association of perioperative dexamethasone, smoking and alcohol abuse with wound complications after laparotomy. Acta Anaesthesiol Scand 2014 Jan 29. doi: 10.1111/aas.12270. [Epub ahead of print]. 42. Straube S, Derry S, Moore RA, Wiffen PJ, McQuay HJ. Single dose oral gabapentin for established acute postoperative pain in adults. Cochrane Database Syst Rev 2010; (5): CD008183. 43. Seib RK, Paul JE. Preoperative gabapentin for postoperative analgesia: a meta-analysis. CMAJ 2006; 53: 461–9. 44. Hurley RW, Cohen SP, Williams KA, Rowlingson AJ, Wu CL. The analgesic effects of perioperative gabapentin on postoperative pain: a meta-analysis. RAPM 2006; 31: 237– 47. 45. Ho KY, Gan TJ, Habib AS. Gabapentin and postoperative pain–a systematic review of randomized controlled trials. Pain 2006; 126: 91–101. 46. Peng PWH, Wijeysundera DN, Li CCF. Use of gabapentin for perioperative pain control – a meta-analysis. Pain Res Manag 2007; 12: 85–92. 47. Tiippana EM, Hamunen K, Kontinen VK, Kalso E. Do surgical patients benefit from perioperative gabapentin/ pregabalin? A systematic review of efficacy and safety. Anesth Analg 2007; 104: 1545–56. 48. Mathiesen O, Moiniche S, Dahl JB. Gabapentin and postoperative pain: a qualitative and quantitative systematic review, with focus on procedure. BMC Anesthesiol 2007; 7: 6. 49. Yu L, Ran B, Li M, Shi Z. Gabapentin and pregabalin in the management of postoperative pain after lumbar spinal surgery: a systematic review and meta-analysis. Spine 2013; 38: 1947–52. 50. Moore RA, Wiffen PJ, Derry S, McQuay HJ. Gabapentin for chronic neuropathic pain and fibromyalgia in adults. Cochrane Database Syst Rev 2011; (3): CD007938.

16

51. Zhang J, Ho KY, Wang Y. Efficacy of pregabalin in acute postoperative pain: a meta-analysis. Br J Anaesth 2011; 106: 454–62. 52. Engelman E, Cateloy F. Efficacy and safety of perioperative pregabalin for post-operative pain: a meta-analysis of randomized-controlled trials. Acta Anaesthesiol Scand 2011; 55: 927–43. 53. Zaccara G, Gangemi P, Perucca P, Specchio L. The adverse event profile of pregabalin: a systematic review and metaanalysis of randomized controlled trials. Epilepsia 2011; 52: 826–36. 54. Hyllested M, Jones S, Pedersen JL, Kehlet H. Comparative effect of paracetamol, nsaids or their combination in postoperative pain management: a qualitative review. Br J Anaesth 2002; 88: 199–214. 55. Romsing J, Moiniche S, Dahl JB. Rectal and parenteral paracetamol, and paracetamol in combination with NSAIDs, for postoperative analgesia. Br J Anaesth 2002; 88: 215– 26. 56. Ong CK, Seymour RA, Lirk P, Merry AF. Combining paracetamol (acetaminophen) with nonsteroidal antiinflammatory drugs: a qualitative systematic review of analgesic efficacy for acute postoperative pain. Anesth Analg 2010; 110: 1170–9. 57. Derry CJ, Derry S, Moore RA. Single dose oral ibuprofen plus paracetamol (acetaminophen) for acute postoperative pain. Cochrane Database Syst Rev 2013; (6): CD010210. 58. de Vries F, Setakis E, van Staa TP. Concomitant use of ibuprofen and paracetamol and the risk of major clinical safety outcomes. Br J Clin Pharmacol 2010; 70: 429–38. 59. Graham GG, Davies MJ, Day RO, Mohamudally A, Scott KF. The modern pharmacology of paracetamol: therapeutic actions, mechanism of action, metabolism, toxicity and recent pharmacological findings. Inflammopharmacology 2013; 21: 201–32. 60. Fitzgerald GA. Coxibs and cardiovascular disease. N Engl J Med 2004; 351: 1709–11. 61. Fosbol EL, Folke F, Jacobsen S, Rasmussen JN, Sorensen R, Schramm TK, Andersen SS, Rasmussen S, Poulsen HE, Kober L, Torp-Pedersen C, Gislason GH. Cause-specific cardiovascular risk associated with nonsteroidal antiinflammatory drugs among healthy individuals. Circ Cardiovasc Qual Outcomes 2010; 3: 395–405. 62. Fosbol EL, Gislason GH, Jacobsen S, Folke F, Hansen ML, Schramm TK, Sorensen R, Rasmussen JN, Andersen SS, Abildstrom SZ, Traerup J, Poulsen HE, Rasmussen S, Kober L, Torp-Pedersen C. Risk of myocardial infarction and death associated with the use of nonsteroidal anti-inflammatory drugs (NSAIDs) among healthy individuals: a nationwide cohort study. Clin Pharmacol Ther 2009; 85: 190–7. 63. Ofman JJ, MacLean CH, Straus WL, Morton SC, Berger ML, Roth EA, Shekelle P. A metaanalysis of severe upper gastrointestinal complications of nonsteroidal antiinflammatory drugs. J Rheumatol 2002; 29: 804–12. 64. Bhangu A, Singh P, Fitzgerald JE, Slesser A, Tekkis P. Postoperative nonsteroidal anti-inflammatory drugs and risk of anastomotic leak: meta-analysis of clinical and experimental studies. World J Surg 2014 Mar 31. [Epub ahead of print]. 65. Devereaux PJ, Mrkobrada M, Sessler DI, Leslie K, Alonso-Coello P, Kurz A, Villar JC, Sigamani A, Biccard BM, Meyhoff CS, Parlow JL, Guyatt G, Robinson A, Garg AX, Rodseth RN, Botto F, Buse GL, Xavier D, Chan MT, Tiboni M, Cook D, Kumar PA, Forget P, Malaga G, Fleischmann E, Amir M, Eikelboom J, Mizera R, Torres D, Wang CY, Vanhelder T, Paniagua P, Berwanger O, Srinathan S, Graham M, Pasin L, Le Manach Y, Gao P, Pogue J, Whitlock R, Lamy A, Kearon C, Baigent C, Chow C, Pettit S, Chrolavicius S,

Adverse effects of multimodal analgesia

66.

67.

68.

69.

Yusuf S. Aspirin in patients undergoing noncardiac surgery. N Eng J Med 2014; 370: 1494–503. Rosseland LA, Romundstad L. Multimodal pain management with glucocorticoids: killing two birds with one stone or shooting sparrows with cannons? Acta Anaesthesiol Scand 2013; 57: 819–22. Stubhaug A, Romundstad L, Kaasa T, Breivik H. Methylprednisolone and ketorolac rapidly reduce hyperalgesia around a skin burn injury and increase pressure pain thresholds. Acta Anaesthesiol Scand 2007; 51: 1138–46. Lunn TH, Kristensen BB, Andersen LO, Husted H, Otte KS, Gaarn-Larsen L, Kehlet H. Effect of high-dose preoperative methylprednisolone on pain and recovery after total knee arthroplasty: a randomized, placebo-controlled trial. Br J Anaesth 2011; 106: 230–8. Bisgaard T, Klarskov B, Kehlet H, Rosenberg J. Preoperative dexamethasone improves surgical outcome after laparoscopic cholecystectomy: a randomized double-blind placebo-controlled trial. Ann Surg 2003; 238: 651–60.

70. Turan A, Sessler DI. Steroids to ameliorate postoperative pain. Anesthesiology 2011; 115: 457–9. 71. Dhatariya K. II. Does dexamethasone-induced hyperglycaemia contribute to postoperative morbidity and mortality? Br J Anaesth 2013; 110: 674–5. 72. Horton R. Vioxx, the implosion of Merck, and aftershocks at the FDA. Lancet 2004; 364: 1995–6. 73. Jüni P, Nartey L, Reichenbach S, Sterchi R, Dieppe PA, Egger M. Risk of cardiovascular events and rofecoxib: cumulative meta-analysis. Lancet 2004; 364: 2021–9. Address: Ole Mathiesen Section of Acute Pain Management Department of Anaesthesiology 4231 Centre of Head and Orthopaedics, Rigshospitalet Blegdamsvej 9 DK-2100 Copenhagen Ø Denmark e-mail: [email protected]

17