Portal de Periódicos da CAPES

Controversy of non-steroidal anti-inflammatory drugs and intracranial surgery: et ne nos inducas in tentationem?

2011; Elsevier BV; Volume: 107; Issue: 3 Linguagem: Inglês

10.1093/bja/aer230

1471-6771

Katharine Kelly, Marc Janssens, J. Ross, E. H. Horn,

Cardiac, Anesthesia and Surgical Outcomes

The paper by Williams and colleagues1Williams DL Pemberton E Leslie K Effect of intravenous parecoxib on post-craniotomy pain.Br J Anaesth. 2011; 107: 398-403Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar in this issue of the British Journal of Anaesthesia touches on the contentious issue of the use of non-steroidal anti-inflammatory drugs (NSAIDs) and intracranial surgery. They found that 40 mg of i.v. parecoxib administered at dural closure in patients undergoing supratentorial neurosurgical procedures had no effect on morphine consumption or pain intensity compared with placebo. There were also no differences in side-effects such as postoperative nausea and vomiting (PONV) and sedation scores. While recording no major morbidity, one issue not specifically mentioned was whether there was any difference in blood loss or conversely thrombotic events between the two groups. They powered their study on the primary outcome of 24 h morphine consumption and not specifically to look at bleeding or thrombotic episodes. NSAIDs are undoubtedly useful analgesics. However, like any drug, they are not free from side-effects.2Rao PNP Knaus EE Evolution of nonsteroidal anti-inflammatory drugs (NSAIDs): cyclooxygenase (COX) inhibition and beyond.J Pharm Pharmaceut Sci. 2008; 11: 81s-110sCrossref PubMed Google Scholar The particular worry with NSAIDs and intracranial procedures is the perceived effect on haemostasis, particularly platelet function. While there are a number of papers looking at the effects of bleeding in non-neurosurgical patients, there is a relative dearth of information in the anaesthetic literature looking at the effects of blood loss specifically in neurosurgical groups. Whereas studies in non-neurosurgical cohorts showing an excess of blood loss3Robinson C Christie J Malcolm-Smith N Nonsteroidal antiinflammatory drugs, perioperative blood loss, and transfusion requirements in elective hip arthroplasty.J Arthroplasty. 1993; 8: 607-610doi:10.1016/0883-5403(93)90007-QAbstract Full Text PDF PubMed Scopus (75) Google Scholar 4Maund E McDaid C Rice S Wright K Jenkins B Woolacott N Paracetamol and selective and non-selective non-steroidal anti-inflammatory drugs for the reduction in morphine-related side-effects after major surgery: a systematic review.Br J Anaesth. 2011; 106: 292-297doi:10.1093/bja/aeq406Abstract Full Text Full Text PDF PubMed Scopus (336) Google Scholar may not necessarily have devastating effects, if that blood loss is within the confines of the skull and particularly in the posterior fossa, that excess may well prove fatal or increase morbidity.5Palmer JD Sparrow OC Iannotti F Postoperative hematoma: a 5-year survey and identification of avoidable risk factors.Neurosurgery. 1994; 35 (discussion 64–5): 1061-1064doi:10.1227/00006123-199412000-00007Crossref PubMed Scopus (273) Google Scholar NSAIDs exert their analgesic properties by inhibiting arachidonic acid-derived prostaglandins (PGs) generated via the cyclooxygenase (COX) pathway.2Rao PNP Knaus EE Evolution of nonsteroidal anti-inflammatory drugs (NSAIDs): cyclooxygenase (COX) inhibition and beyond.J Pharm Pharmaceut Sci. 2008; 11: 81s-110sCrossref PubMed Google Scholar In the early 1990s, a differentiation was made between two isoforms of COX; a constitutive form producing 'anti-inflammatory' PGs (COX-1) and an inducible pro-inflammatory isoform subsequently identified as COX-2.2Rao PNP Knaus EE Evolution of nonsteroidal anti-inflammatory drugs (NSAIDs): cyclooxygenase (COX) inhibition and beyond.J Pharm Pharmaceut Sci. 2008; 11: 81s-110sCrossref PubMed Google Scholar Arachidonic acid products also exert an effect on haemostasis by being 'prothrombotic' (thromboxane A2 from platelets) or 'anti-thrombotic' (Prostacyclin, PGI2, from the endothelial cells). One can consider this as a balance between these two systems.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar After vascular injury, endothelial cells are damaged and this balance is impaired. Platelets adhere to the damaged vessel wall, due to a number of factors including von Willebrand's factor and complex interactions of adhesion molecules, but also thromboxane A2, synthesized in activated platelets. There is a relative reduction of PGI2 synthesis from injured endothelial cells.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar We should stress the difference here between aspirin and other NSAIDs. Aspirin strictly speaking is 'non-steroidal', 'anti-inflammatory', and 'analgesic', but in common anaesthetic parlance, 'Non-steroidal anti-inflammatory drugs' (NSAIDs) generally refer to drugs such as ibuprofen, naproxen, or ketorolac (non-selective COX-1 and COX-2 inhibitors) or the more selective COX-2 inhibitors such as parecoxib. The main subject of this editorial is the NSAIDs, but it is useful to mention aspirin here to illustrate how it differs from other NSAIDs. Aspirin covalently and irreversibly binds to platelet COX. As platelets are anucleic, platelet COX activity cannot recover until new platelets are produced in the bone marrow. Aspirin is also thought to have a greater effect on the platelet COX producing thromboxane A2, than the endothelial cell COX producing PGI2. The entire cohort of platelets is replaced approximately every 10 days. Thus, some clinicians will stop aspirin anywhere between 7 and 14 days before surgery, by when all platelets exposed to aspirin will have been replaced.3Robinson C Christie J Malcolm-Smith N Nonsteroidal antiinflammatory drugs, perioperative blood loss, and transfusion requirements in elective hip arthroplasty.J Arthroplasty. 1993; 8: 607-610doi:10.1016/0883-5403(93)90007-QAbstract Full Text PDF PubMed Scopus (75) Google Scholar 6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar This has to be balanced against the risks of stopping the aspirin, thus depriving the patient of the therapeutic effects for treatment of the condition that aspirin was commenced to treat. Usually this is to prevent cardiac or cerebrovascular disease. This may require careful dialogue between several specialties including the neurosurgeon, cardiologist (or neurologist or stroke physician), haematologist and anaesthetist. Non-aspirin NSAIDs reversibly inhibit COX, by a varying duration and extent depending on the individual agent.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar Ex vivo tests of platelet aggregation, such as second-wave aggregation to ADP or epinephrine after a single dose of drug, show prolongation for 3 days after piroxicam, 2 days after naproxen, diclofenac, and indomethacin, and about 1 day after ibuprofen. With chronic administration, the time taken for return of normal platelet function may be considerably longer, due to accumulation of products such as the S(+) enantiomer of ibuprofen in adipose tissue.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar Evidence of clinical blood loss at non-neurosurgical operations is conflicting. For indomethacin, this has been variably reported as causing increased bleeding at abdominal surgery and hysterectomy, while not causing increased blood loss in emergency surgery to the lower limb.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar General advice is that non-aspirin NSAIDs should be stopped before surgery on sites requiring critical haemostasis or in patients with a pre-existing coagulopathy.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar This period of time is probably shorter than that required for aspirin as these other NSAIDs have a reversible effect. The difficulty is determining the precise advice for an individual patient, on a specific NSAID. Should this advice be taken from Palmer and colleagues5Palmer JD Sparrow OC Iannotti F Postoperative hematoma: a 5-year survey and identification of avoidable risk factors.Neurosurgery. 1994; 35 (discussion 64–5): 1061-1064doi:10.1227/00006123-199412000-00007Crossref PubMed Scopus (273) Google Scholar who considered aspirin and other NSAIDs together and advised stopping antiplatelet agents 2 weeks in advance of neurosurgery, or from older papers in fields such as orthopaedics (stop NSAIDs 5–7 days before operation),3Robinson C Christie J Malcolm-Smith N Nonsteroidal antiinflammatory drugs, perioperative blood loss, and transfusion requirements in elective hip arthroplasty.J Arthroplasty. 1993; 8: 607-610doi:10.1016/0883-5403(93)90007-QAbstract Full Text PDF PubMed Scopus (75) Google Scholar or from more recent papers which suggest a substantially shorter period of time for non-aspirin NSAIDs?6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar While doubt exists, some would pragmatically suggest a 2 week omission of non-aspirin NSAIDs. This is a decision for all specialities involved with the individual patient. These issues have been reviewed.6Shafer AI Effects of nonsteroidal antiinflammatory drugs on platelet function and systemic hemostasis.J Clin Pharmacol. 1995; 35: 209-219Crossref PubMed Scopus (299) Google Scholar The use of NSAIDs after craniotomy is inconsistent. A recurrent concern is postoperative haematoma.7Sebastian J Hunt K United Kingdom and Ireland survey of the use of non-steroidal anti-inflammatory drugs after intracranial surgery.J Neurosurg Anesthesiol. 2006; 18: 267doi:10.1097/00008506-200610000-00013Crossref PubMed Scopus (7) Google Scholar In 1995, the overall incidence of post-craniotomy haematoma was reported as 2.2%, and that 88% of patients developing a haematoma did so before 6 h.8Taylor WA Thomas NW Wellings JA Bell BA Timing of postoperative intracranial hematoma development and implications for the best use of neurosurgical intensive care.J Neurosurg. 1995; 82: 48-50doi:10.3171/jns.1995.82.1.0048Crossref PubMed Scopus (125) Google Scholar Thus, some colleagues justify postoperative use of NSAIDs after 6 h.7Sebastian J Hunt K United Kingdom and Ireland survey of the use of non-steroidal anti-inflammatory drugs after intracranial surgery.J Neurosurg Anesthesiol. 2006; 18: 267doi:10.1097/00008506-200610000-00013Crossref PubMed Scopus (7) Google Scholar This seems flawed, as the corollary of this paper is that 12% will still develop haematoma after 6 h. A dichotomous distribution of either before 6 h or after 24 h was actually reported.8Taylor WA Thomas NW Wellings JA Bell BA Timing of postoperative intracranial hematoma development and implications for the best use of neurosurgical intensive care.J Neurosurg. 1995; 82: 48-50doi:10.3171/jns.1995.82.1.0048Crossref PubMed Scopus (125) Google Scholar If in the interim, an NSAID has been administered, it will be hard to argue that this did not in some way contribute to the expanding haematoma. A survey of British and Irish practice of prescribing NSAIDs after craniotomy found that 64% of respondents (23 centres) did not routinely prescribe these agents.7Sebastian J Hunt K United Kingdom and Ireland survey of the use of non-steroidal anti-inflammatory drugs after intracranial surgery.J Neurosurg Anesthesiol. 2006; 18: 267doi:10.1097/00008506-200610000-00013Crossref PubMed Scopus (7) Google Scholar Oddly in the light of the earlier paper,8Taylor WA Thomas NW Wellings JA Bell BA Timing of postoperative intracranial hematoma development and implications for the best use of neurosurgical intensive care.J Neurosurg. 1995; 82: 48-50doi:10.3171/jns.1995.82.1.0048Crossref PubMed Scopus (125) Google Scholar they also found that of the 13 centres who did prescribe, 54% administered NSAIDs either perioperatively or immediately after operation, with three centres waiting 6 h, 1 centre 12 h, and two centres 24 h.7Sebastian J Hunt K United Kingdom and Ireland survey of the use of non-steroidal anti-inflammatory drugs after intracranial surgery.J Neurosurg Anesthesiol. 2006; 18: 267doi:10.1097/00008506-200610000-00013Crossref PubMed Scopus (7) Google Scholar A survey in 2009 of all 31 adult neurosurgical centres in the UK found that only 52% (16 units) used NSAIDs after craniotomy, and only three of 31 units administered NSAIDs regularly. The average time to first dose was 24 h, and the range was from immediately postoperative to after 72 h. Of these 16 units, 14 used diclofenac and two used ibuprofen, that is, non-selective COX inhibitors. None used COX-2 inhibitors. In the units that used NSAIDs non-routinely, use was governed by the individual surgeon's choice.9Kotak D Cheserem B Solth A A survey of post-craniotomy analgesia in British neurosurgical centres: time for perceptions and prescribing to change?.Br J Neurosurg. 2009; 23: 538-542doi:10.1080/02688690903100595Crossref PubMed Scopus (30) Google Scholar The incidence of haematoma may depend on the exact procedure undertaken. A 5 yr review of postoperative haematoma in 6668 neurosurgical procedures in one unit quoted an overall incidence of 1.1%.5Palmer JD Sparrow OC Iannotti F Postoperative hematoma: a 5-year survey and identification of avoidable risk factors.Neurosurgery. 1994; 35 (discussion 64–5): 1061-1064doi:10.1227/00006123-199412000-00007Crossref PubMed Scopus (273) Google Scholar This survey included all neurosurgical procedures not just intracranial. However, the incidence varied from 0% in their centre for lumbar surgery up to 6.2% after surgery for meningioma. Trauma accounted for 3.7%, aneurysm surgery 2.6%, and intrinsic supratentorial tumours 2.2%. Parecoxib is a COX-2 inhibitor, a class of drug said to be free from inhibiting platelet function. However, there is now a worry that COX-2 inhibitors may go to the other extreme and promote thrombosis.10McGettigan P Henry D Cardiovascular risk and inhibition of cyclooxygenase. A systematic review of the observational studies of selective and nonselective inhibitors of cyclooxygenase 2.J Am Med Assoc. 2006; 296: 1633-1644doi:10.1001/jama.296.13.jrv60011Crossref PubMed Scopus (980) Google Scholar This led to the withdrawal of the COX-2 inhibitor rofecoxib in 2004.11Bresalier RS Sandler RS Quan H et al.Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial.N Engl J Med. 2005; 352: 1092-1102doi:10.1056/NEJMoa050493Crossref PubMed Scopus (2293) Google Scholar This risk was initially thought to be present only with prolonged administration, but the time period seems to be shortening. Early papers described risks after 18 months,11Bresalier RS Sandler RS Quan H et al.Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial.N Engl J Med. 2005; 352: 1092-1102doi:10.1056/NEJMoa050493Crossref PubMed Scopus (2293) Google Scholar and a systematic review of observational studies of selective and non-selective inhibitors of COX-2 concluded that this adverse risk could be observed after 30 days of COX-2 administration.10McGettigan P Henry D Cardiovascular risk and inhibition of cyclooxygenase. A systematic review of the observational studies of selective and nonselective inhibitors of cyclooxygenase 2.J Am Med Assoc. 2006; 296: 1633-1644doi:10.1001/jama.296.13.jrv60011Crossref PubMed Scopus (980) Google Scholar Myocardial infarction, cardiac arrest, stroke, and pulmonary embolism were all more common in patients given either parecoxib or valdecoxib (another COX-2 inhibitor) compared with placebo after cardiac surgery, after only a 10 day course.12Nussmeier NA Whelton AA Brown MT et al.Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery.N Engl J Med. 2005; 352: 1081-1091doi:10.1056/NEJMoa050330Crossref PubMed Scopus (963) Google Scholar Could this period be even shorter? There is recent evidence that the thrombotic tendency may not be confined to COX-2 inhibitors but also may affect non-selective NSAIDs such as diclofenac or ibuprofen.10McGettigan P Henry D Cardiovascular risk and inhibition of cyclooxygenase. A systematic review of the observational studies of selective and nonselective inhibitors of cyclooxygenase 2.J Am Med Assoc. 2006; 296: 1633-1644doi:10.1001/jama.296.13.jrv60011Crossref PubMed Scopus (980) Google Scholar 13Trelle S Reichenbach S Wandel S et al.Cardiovascular safety of non-steroidal anti-inflammatory drugs: network meta-analysis.Br Med J. 2011; 342: c7086doi:10.1136/bmj.c7086Crossref PubMed Scopus (879) Google Scholar Thus, these non-specific NSAIDs have the negative 'double whammy' of inhibiting platelet function in some patients while also (and counter-intuitively) promoting thrombotic tendency in others. It may be that some 'non-selective' NSAIDs such as diclofenac have a greater effect on COX-2 and are more selective than originally thought.10McGettigan P Henry D Cardiovascular risk and inhibition of cyclooxygenase. A systematic review of the observational studies of selective and nonselective inhibitors of cyclooxygenase 2.J Am Med Assoc. 2006; 296: 1633-1644doi:10.1001/jama.296.13.jrv60011Crossref PubMed Scopus (980) Google Scholar This may well reflect the number of interactive mechanisms involving individual drugs and this balance in a specific patient. For intracranial surgery, these questions could be addressed either by systematic reviews or by a randomized, controlled trial (RCT) specifically designed to answer the question of the effects of NSAIDs in intracranial surgery, by an appropriate professional body such as the Neuroanaesthesia Society, in the UK. There have been calls for large randomized safety and efficacy trials to answer this particular problem.14Umamaheswara Rao GS Gelb AW To use or not to use: the dilemma of NSAIDs and craniotomy.Eur J Anaesthesiol. 2009; 26: 625-626doi:10.1097/EJA.0b013e32832a21adCrossref PubMed Scopus (17) Google Scholar However, this would require a definition of what would be considered an acceptable level of potential adverse events in the treatment group. Williams and colleagues1Williams DL Pemberton E Leslie K Effect of intravenous parecoxib on post-craniotomy pain.Br J Anaesth. 2011; 107: 398-403Abstract Full Text Full Text PDF PubMed Scopus (33) Google Scholar suggest that systematic reviews may be the only way to establish these and other risks. This would also be an opportunity to rationalize analgesic regimens after intracranial surgery. We live in an era of evidence-based medicine, yet there is still no consensus of postoperative analgesia in intracranial surgery. While this allows each unit to tailor analgesic use to its own requirements, it also allows less logical practice. A survey in 1995 found that 90% of neurosurgical centres replying to a questionnaire used codeine as the mainstay of post-craniotomy analgesia, despite 50% of respondents considering it to be a poor analgesic.15Stoneham MD Walters FJ Post-operative analgesia for craniotomy patients: current attitudes among neuroanaesthetists.Eur J Anaesthesiol. 1995; 12: 571-575PubMed Google Scholar Codeine persists as the mainstay of opioid-based analgesia in 70% of UK neurosurgical centres who responded to a survey in 2009,9Kotak D Cheserem B Solth A A survey of post-craniotomy analgesia in British neurosurgical centres: time for perceptions and prescribing to change?.Br J Neurosurg. 2009; 23: 538-542doi:10.1080/02688690903100595Crossref PubMed Scopus (30) Google Scholar perhaps due to the historical perception that there is less interference with neurological signs. This is despite evidence that in doses used in clinical practice, morphine causes no more sedation than codeine, but morphine gave more predictable and prolonged analgesia.16Goldsack C Scuplak SM Smith M A double-blind comparison of codeine and morphine for postoperative analgesia following intracranial surgery.Anaesthesia. 1996; 51: 1029-1032doi:10.1111/j.1365-2044.1996.tb14997.xCrossref PubMed Scopus (66) Google Scholar Codeine also partially undergoes enzymatic conversion to morphine for some of its analgesic effect. A number of patients lack the ability to do this, thus limiting its effect.16Goldsack C Scuplak SM Smith M A double-blind comparison of codeine and morphine for postoperative analgesia following intracranial surgery.Anaesthesia. 1996; 51: 1029-1032doi:10.1111/j.1365-2044.1996.tb14997.xCrossref PubMed Scopus (66) Google Scholar Other methods have included paracetamol, other opioids, tramadol (usually as a third- or fourth-line analgesic), and NSAIDs,9Kotak D Cheserem B Solth A A survey of post-craniotomy analgesia in British neurosurgical centres: time for perceptions and prescribing to change?.Br J Neurosurg. 2009; 23: 538-542doi:10.1080/02688690903100595Crossref PubMed Scopus (30) Google Scholar but all drugs have advantages, disadvantages, and side-effects. Paracetamol used in appropriate doses is a 'safe well-tolerated drug with proven efficacy'.17Oscier C Milner Q Peri-operative use of paracetamol.Anaesthesia. 2009; 64: 65-72doi:10.1111/j.1365-2044.2008.05674.xCrossref PubMed Scopus (84) Google Scholar However, it is a weaker analgesic than others as described by numbers needed to treat in some models of moderate-to-severe pain.17Oscier C Milner Q Peri-operative use of paracetamol.Anaesthesia. 2009; 64: 65-72doi:10.1111/j.1365-2044.2008.05674.xCrossref PubMed Scopus (84) Google Scholar Stronger analgesics, such as opioids, may cause respiratory depression and also nausea and vomiting, and the fear of this has led neuroanaesthetists to shy away from their use post-craniotomy.15Stoneham MD Walters FJ Post-operative analgesia for craniotomy patients: current attitudes among neuroanaesthetists.Eur J Anaesthesiol. 1995; 12: 571-575PubMed Google Scholar Williams and colleagues comment that they did not include prophylactic antiemetics, although dexamethasone was given, presumably to reduce peri-tumour oedema. If there is an intent to reduce PONV, perhaps a more effective way would be an evidence-based policy of antiemetic measures.18Tramer MR Systematic reviews in PONV therapy.in: Evidence Based Resource in Anaesthesia and Analgesia. BMJ Books, London2000: 157-178Google Scholar Tramadol is associated with convulsions which is surely a major disadvantage in intracranial surgery.9Kotak D Cheserem B Solth A A survey of post-craniotomy analgesia in British neurosurgical centres: time for perceptions and prescribing to change?.Br J Neurosurg. 2009; 23: 538-542doi:10.1080/02688690903100595Crossref PubMed Scopus (30) Google Scholar 19Kahn LH Alderfer RJ Graham DJ Seizures reported with tramadol.J Am Med Assoc. 1997; 278: 1661doi:10.1001/jama.278.20.1661bCrossref PubMed Scopus (97) Google Scholar Aside from their use of a COX-2 inhibitor, Williams and colleagues' other methods of scalp infiltration with local anaesthetic, perioperative i.v. paracetamol, and in appropriate cases postoperative patient-controlled anaglesia opioid (usually morphine), is essentially similar to the one used in our unit in Edinburgh. Proponents of NSAID administration in the perioperative or immediate postoperative period should note the variable requirements for analgesia after craniotomy.20De Benedittis G Lorenzetti A Migliore M Spagnoli D Tiberio F Villani R Postoperative pain in neurosurgery: a pilot study in brain surgery.Neurosurgery. 1996; 38: 466-470PubMed Google Scholar NSAIDs may not be necessary. Williams and colleagues found that there was no significant difference between the control group and the group who had received parecoxib. They achieved 'excellent' or 'very good' pain scores in 74% and 78%, respectively. Although a larger dose of dexamethasone was given to the control group than the parecoxib group, and dexamethasone may have analgesic properties,21Hockey B Leslie K Williams D Dexamethasone for intracranial neurosurgery and anaesthesia.J Clin Neurosci. 2009; 16: 1389-1393doi:10.1016/j.jocn.2009.03.007Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar the small difference from control mirrored the limited effect seen in a previous study on parecoxib after craniotomy reported in this journal in 2009.22Jones S Cormack J Murphy M Scott D Parecoxib for analgesia after craniotomy.Br J Anaesth. 2009; 102: 76-79doi:10.1093/bja/aen318Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar Thus, giving parecoxib to all patients, some who would not have required it, may subject these patients to the risk of adverse events for no gain. Finally, although COX-2 inhibitors may protect against bleeding and also gastric irritation, there are side-effects common to both selective and non-selective NSAIDS, such as renal adverse events.23Anonymous Cardiovascular safety of NSAIDs and selective COX-inhibitors.Curr Probl Pharmacovigilance. 2006; 31: 7Google Scholar In summary, when considering the routine administration of NSAIDs perioperatively for intracranial procedures, the following points should be borne in mind. Concerns about NSAIDs effect on haemostasis round the time of craniotomy persist, many peers do not prescribe these agents, the patient may not have sufficient pain to warrant their use, and finally, NSAIDs may also cause other adverse events such as renal failure. In view of these concerns, and in the absence of an adequately powered RCT of the safety of NSAIDs in this patient group, should we not take a pragmatic approach? That is, on current levels of evidence, NSAIDs should not be administered routinely in close proximity to craniotomy; that a period of at least 24 h should elapse before administering these agents and only then if satisfactory analgesia is not being obtained. Potentially, there may be more gain after operation in operations which are more painful, such as posterior fossa procedures.20De Benedittis G Lorenzetti A Migliore M Spagnoli D Tiberio F Villani R Postoperative pain in neurosurgery: a pilot study in brain surgery.Neurosurgery. 1996; 38: 466-470PubMed Google Scholar However, infratentorial haematomas are associated with a worse outcome,5Palmer JD Sparrow OC Iannotti F Postoperative hematoma: a 5-year survey and identification of avoidable risk factors.Neurosurgery. 1994; 35 (discussion 64–5): 1061-1064doi:10.1227/00006123-199412000-00007Crossref PubMed Scopus (273) Google Scholar and we must suggest again, on current levels of evidence, NSAIDs should only be used after other analgesics are inadequate, and after a sufficient period of time has elapsed since the procedure, and also after discussion with the relevant surgeon. Until there is good evidence of efficacy and safety, many neurosurgeons would take the view that NSAIDs' use after cranial surgery is contra-indicated in general, and absolutely contra-indicated in the postoperative management of 'high risk' cases such as tumours, vascular cases, and trauma. Returning to the subtitle of this editorial (et ne nos inducas in tentationem), we should not be led into the temptation of a blanket administration of NSAIDs at intracranial surgery or its immediate aftermath. None declared.