Page 122 Acute Pain Management

The combination of paracetamol and nsNSAID was clearly more effective than paracetamol
alone, but evidence for superiority relative to the nsNSAID alone was more limited and of
uncertain clinical significance (Hyllested et al, 2002 Level I; Romsing et al, 2002 Level I).
Adverse effects
NsNSAID side effects are more common with long‐term use — and there are concerns relating
to prothrombotic effects. In the perioperative period the main concerns are renal impairment,
interference with platelet function, wound and bone healing, and peptic ulceration or
bronchospasm in individuals at risk. Certain risks are accentuated in the perioperative period
because of haemodynamic disturbances, fluid shifts, activation of the neurohumoral stress
response and deficient enteral feeding. In general, the risk and severity of nsNSAID‐associated
side effects is increased in elderly people (Pilotto et al, 2003; Juhlin et al, 2005).
Renal function
Renal prostaglandins regulate tubular electrolyte handling, modulate the actions of renal
hormones, and maintain renal blood flow and glomerular filtration rate in the presence of
circulating vasoconstrictors. The adverse renal effects of chronic nsNSAID use are common and
well‐recognised. In some clinical conditions, including hypovolaemia, dehydration and major
surgery, high circulating concentrations of the vasoconstrictors angiotensin II, noradrenaline
and vasopressin increase production of intrarenal vasodilators including prostacyclin —
CHAPTER 4 maintenance of renal function may then depend on prostaglandin synthesis and thus can be
sensitive to brief nsNSAID administration.
In patients with normal preoperative renal function, nsNSAIDs caused a clinically insignificant
and transient decrease in creatinine clearance the first day after surgery, and there were no
differences between patients given diclofenac, ketorolac, indomethacin (indometacin) or
ketoprofen (Lee A et al, 2007 Level I). The risk of adverse renal effects of nsNSAIDs and coxibs is
increased in the presence of factors such as pre‐existing renal impairment, hypovolaemia,
hypotension, use of other nephrotoxic agents and angiotensin‐converting enzyme (ACE)
inhibitors (RCA, 1998 Level IV).
With proper selection and monitoring, the incidence of NSAID‐induced perioperative renal
impairment is low and NSAIDs need not be withheld in patients with normal preoperative
renal function (Lee A et al, 2007 Level I).
Platelet function
NsNSAIDs inhibit platelet function. In meta‐analyses of tonsillectomy in both adult and
paediatric patients, nsNSAIDs were found to increase the risk of reoperation for bleeding (NNH
29 to 60) (Marret et al, 2003 Level I; Moiniche et al, 2003 Level I) but surgical blood loss was not
significantly increased (Moiniche et al, 2003 Level I) (see also Sections 9.6.7 and 10.5). Looking at
studies in children only, there was no increase in the risk of reoperation for bleeding after
tonsillectomy (Cardwell et al, 2005 Level I). Aspirin, which irreversibly inhibits platelet
aggregation, increased the risk of post‐tonsillectomy haemorrhage (Krishna et al, 2003 Level I).
After a variety of different operations, the use of nsNSAIDs showed a significant increase in
risk of severe bleeding from 0 to 1.7% compared with placebo (NNH 59) (Elia et al, 2005 Level I).
This was also found in the large HIPAID study after hip replacement, where the ibuprofen
group had a significantly increased risk of major bleeding complications (OR 2.1) (Fransen et al,
2006 Level II). After gynaecological or breast surgery, use of diclofenac led to more blood loss
than rofecoxib (Hegi et al, 2004 Level II). After otorhinolaryngological surgery in an outpatient
setting, tenoxicam increased bleeding at the surgical site (Merry et al, 2004 Level II).

74 Acute Pain Management: Scientific Evidence