The most common method of postoperative pain relief is the traditional use of 'on demand' intramuscular opioid injections. Better pain relief can be obtained with newer techniques such as epidural opioids and patient controlled analgesia. However, some techniques such as pre-emptive analgesia have not been fully evaluated and others, such as infiltrating local anaesthetic into wounds, may not reduce the patient's analgesic requirements. Patients should be counselled before surgery because an explanation of what to expect, and other simple measures, may reduce their analgesic needs.


'For all the happiness mankind can gain is not in pleasure but in rest from pain' - John Dryden.

Although postoperative pain is arguably the most common clinical problem in our hospitals, it is often dismissed with an order for intermittent intramuscular opiate injections to be given at the discretion of an overworked nursing staff. This generally results in patients waiting for pain relief, then a period of relief and perhaps drowsiness, and then the cycle is repeated. With this method, pain relief is only satisfactory (i.e. adequate relief without unwanted sedation) for about one-third of the time.1

At least 3 factors contribute to the persistence of such poor practice. There is a community attitude that to betray distress from pain is weak. This tends to draw a negative response from staff, sometimes quite startlingly, when ethnic differences in attitudes to pain are evident. The second factor is medical and nursing training which emphasises the adverse effects of therapy. This sometimes results in an excessive apprehension about possible adverse effects. While opioids can cause respiratory depression and dependence, they have a wide therapeutic ratio and a low risk of dependence when given for pain relief. Finally, the realities of cost containment and hospital staffing limit our ability to provide what we know to be possible. With our knowledge and the technology which is now generally available in Australian hospitals, we can and should do better than this.

Good postoperative analgesic management probably carries benefits other than increased patient comfort. The magnitude of the neuro-endocrine stress response,2 postoperative pulmonary complications and the incidence of myocardial ischaemia can be decreased. Early mobilisation can be achieved and the patient can be discharged from hospital sooner.

Pain mechanisms

The sensory nervous system can distinguish innocuous low intensity stimuli from potentially noxious high intensity stimuli. The former activate low threshold sensors, and the impulse is conducted along Aß fibres. The signals generated at high threshold nociceptors, communicate with the central nervous system via Aδ and C fibres. The conduction of impulses through the spinal cord and brain stem to the thalamus is the simple sensory aspect of pain. The processing of this physiological signal into the percept of pain and the individual's pain behaviour are under the influence of psychological factors. For optimal pain management, all these mechanisms need to be taken into account.

Non-pharmacological methods of pain control

A careful explanation of what the patient can expect to experience after the operation can significantly reduce morphine requirements. Other simple measures which reduce pain include ensuring a comfortable well supported position in bed and splinting the extremities. Relaxation therapy, acupuncture and transcutaneous electrical nerve stimulation (TENS) also reduce analgesic requirements, but by themselves are usually inadequate for postoperative pain. Hypnosis is effective in some, but is very time consuming. Operative neurolysis (e.g. cryotherapy) can be very effective, but recovery of nerve function is slow and neuralgias may follow.

Analgesic drugs

Four classes of drugs are useful in the management of postoperative pain.


This class includes the agonists, morphine, pethidine, fentanyl and methadone, and the partial agonists, buprenorphine and pentazocine. Pentazocine has some antagonist properties and can antagonise the analgesia of the agonist drugs and increase pain if given after morphine or pethidine. Buprenorphine has no significant antagonist activity.

Oxycodone is intermediate in efficacy between the full agonists and paracetamol and the non-steroidal anti-inflammatory drugs (NSAIDs). The place of codeine is unclear. About 10% of an orally administered dose of codeine is converted to morphine, but whether it also has significant analgesic activity of its own is uncertain.

The adverse effects of these drugs include ventilatory depression, depression of respiratory guarding reflexes, nausea and vomiting, drowsiness and urinary retention. Pentazocine also tends to cause dysphoria. Doses need to be reduced in the elderly. In renal failure, fentanyl is the opioid of choice. The actions of all other opioids may be prolonged.

Paracetamol and NSAIDs

The non-steroidal anti-inflammatory drugs (NSAIDs) and paracetamol are adequate after minor to moderate surgery and may reduce opioid requirements after major surgery.

The NSAIDs can cause or exacerbate gastrointestinal upsets, peptic ulcers, platelet dysfunction, renal impairment and asthma. Most of these drugs are available only for oral administration or as suppositories, but the adverse effects can occur independently of the route of administration. The platelet inhibition following acetylsalicylic acid is irreversible and therefore the production of fresh platelets is needed to restore haemostasis. For this reason, it is no longer used as a perioperative analgesic. Ketorolac is an effective analgesic of this class and is available for parenteral as well as oral administration. However, like all NSAIDs, it can cause renal dysfunction, especially in the aged and in the presence of pre-existing renal damage. Some cases of renal failure have occurred in previously healthy people who received ketorolac for postoperative analgesia. Recently, new recommendations for dosage have been issued.

Paracetamol does not appear to cause gastrointestinal complications, platelet dysfunction, renal problems or asthma. It is, however, not a powerful analgesic and in overdose can cause hepatic failure.

Nitrous oxide

The old 'laughing gas' is available with 50% oxygen in premixed cylinders. It is a potent analgesic, probably superior to the opioids in the control of acute somatic pain such as when drain tubes are removed or dressings changed. Of course, it must be breathed for a sufficient time (at least 5 minutes) to establish analgesia. It is usually self-administered.

Local anaesthetics

Although there are other local anaesthetics available, lignocaine and bupivacaine represent a choice of rapid onset and short duration (lignocaine), or longer duration and slower onset.

Other routes

i. Sublingual

Buprenorphine sublingual tablets are effective analgesics. It is uncertain whether buprenorphine's effect is due to sublingual absorption or being swallowed in the saliva.

ii. Transdermal

Fentanyl patches for transdermal administration are being trialed. Absorption is slow, but appears effective. Patches may be useful in paediatrics, but in adults they will probably be used more in the management of chronic pain.

Pre-emptive analgesia

Trauma or inflammation results in sensitisation of pain pathways. Locally, mediators sensitise the Aδ and C fibres so that low intensity stimuli cause pain sensation. Centrally, sensory bombardment results in changes such that input by normal low threshold Aß fibres is sensed as pain. Pre-emptive analgesia4 is based on the hypothesis that these changes can be prevented and that this will result in better analgesia. Many clinical trials using varying protocols have usually (but not always) reported an advantage for pre-emptive techniques. Most reports suggest that giving a local anaesthetic before surgery (locally, regionally, by spinal or epidural routes), or the administration of NSAIDs or opiates, results in reduced pain and/or reduced analgesic requirements postoperatively.

Local preoperative infiltration analgesia appears to be more effective on postoperative pain than single preoperative spinal or epidural administration, not to be confused with continued spinal or epidural postoperative analgesia. Opioids or NSAIDs given with the premedication or during anaesthesia also reduce postoperative pain.

When combination therapies have been studied, it has not been possible to say whether the effects were additive or synergistic. None of the trials have been large enough to exclude the possibility of increased adverse effects, or to judge any impact on postoperative morbidity. At present, the benefits of pre-emptive analgesia are not established.

Methods of opioid administration


The use of oral analgesia in the postoperative period is limited. Methadone is the only agonist with adequate oral bioavailability, but it is very long-acting and accumulates over several days and is therefore little used. Oral slow absorption preparations of morphine are available, but, because of slow onset, they are more useful in palliative care. Oxycodone is sometimes useful in pain of moderate severity.


Oxycodone suppositories are frequently used to manage chronic pain, but can also be used in the postoperative period. Unfortunately, absorption from the rectum is not always reliable, although an effect for up to 12 hours can sometimes be achieved. Tolerance develops in time.


i. Intramuscular

The traditional 'on demand' intramuscular injection of an opioid is still the most common way to relieve postoperative pain. However, in practice, pain control without excessive sedation is only effective for about one third of the time.

ii. Intravenous

Bolus - Small bolus doses of opioid are a rapid and effective albeit labour intensive way of relieving postoperative pain. This method is especially suitable for pain relief in patients immediately after they emerge from anaesthesia in the recovery room. The rapid attainment of maximal effect makes it safe to transport patients to the ward, without fear of increasing respiratory depression after leaving recovery. A less time-consuming means of analgesia needs to be instituted in the ward.

Continuous infusion - Continuous infusion of an opioid is an effective way of achieving good pain control, but requires careful supervision. After initial loading by bolus doses or a higher rate of infusion, analgesia can be maintained usually by 1-3 mg/hour of morphine or 10-25 mg of pethidine/hour. Supervision is important, as stable plasma concentrations will not be reached for many hours (about 4-12 for morphine and 6-20 for pethidine). If analgesia is insufficient, a similar period will elapse before an increase in the infusion rate restores stable concentrations and so repeated, small loading doses are generally needed. Recovery from excessive sedation is equally slow.

Patient controlled analgesia (PCA) - The patient is able to self-administer small intravenous doses of analgesic at prescribed minimum intervals. Although the equipment was developed for research, it is now widely available and can be programmed to deliver a dose at specified minimum intervals when the patient presses a switch. Most devices also enable the prescriber to limit the maximum dose over a fixed interval (e.g. 4 hours), but it is usually unnecessary to impose this limitation. The quality of pain relief is similar to that achieved with continuous infusions, but over dosage is less common. The feeling of being in control of their treatment is usually preferred by patients, and satisfaction with PCA is usually high. The method has been used in children as young as 4 years and is generally useful in children older than 6 years.

Most PCA devices allow a continuous infusion to be programmed in addition to patient controlled doses. This method may improve analgesia marginally, especially on awakening after sleep, but the incidence of over dosage is significantly increased.

iii. Epidural

Injecting opioids into the epidural space results in prolonged analgesia. Bolus doses of morphine 2-4 mg followed by hourly rates of 0.2-0.4 mg/hour, pethidine 30-50 mg and 2-4 mg/hour or fentanyl 50-100 micrograms and 25-50 micrograms/hour give good postoperative analgesia with most patients free of pain at rest. With morphine and pethidine, the quality of epidural analgesia is generally thought to be better than that achieved by PCA or infusion, but with fentanyl the quality is similar. The diluting volume does not affect efficacy, but there is some evidence that a large volume may cause more respiratory depression. The addition of adrenaline may prolong the effect, and the addition of clonidine may enhance analgesia, although the literature is somewhat contradictory.

Adding a low concentration of bupivacaine to opioid epidural infusions is controversial. Following Caesarean section, the addition of fentanyl to bupivacaine enhances analgesia, but after abdominal or thoracic surgery, the addition of bupivacaine to fentanyl has not resulted in improved analgesia in several studies. However, the addition of 100 micrograms of fentanyl to a 4 mg bolus of morphine increased efficacy more than could be expected from a simple additive effect.

The adverse effects of epidural opioids include pruritus, nausea, vomiting, urinary retention, sedation and respiratory depression. Following the administration of morphine, the incidence of severe respiratory depression, which may be delayed as long as 18 hours, is 0.1-0.5%. This phenomenon is rare with fentanyl, if it occurs at all, but a number of instances of severe respiratory depression have been reported following high (>100 micrograms/hour) doses. This is hardly surprising as drug uptake from the epidural space is more rapid than following intramuscular injection. The risk of delayed respiratory depression following pethidine is not clear, but appears to be less than that following morphine.

Antagonism of opioid-induced respiratory depression

Naloxone is able to reverse opioid-induced sedation and respiratory depression. Caution should be exercised, however, because analgesia is also reversed and sudden severe pain, agitation and severe hypertension can be caused by the injudicious use of the naloxone. Moderate respiratory depression (respiration rate >8/minute) can be treated with the administration of oxygen alone, while more severe depression should be treated with small intravenous bolus doses (80-100 micrograms) of naloxone until the respiration is adequate. The duration of action of naloxone is much shorter than that of morphine, pethidine or fentanyl and therefore continued observation is mandatory. In opioid dependent patients, naloxone is likely to precipitate acute drug withdrawal.

Local anaesthetic techniques


In practice, infiltrating the wound line with local anaesthetic before incision, or after closure, usually delays the need for other analgesia by 3 to 6 hours. This is of significant benefit after minor operations, but after major operations, the total analgesic requirement is only reduced slightly. Although the onset of pain is delayed, it is often reported as more severe than after placebo infiltration.

Intercostal nerve blocks

Intercostal nerve blocks are easy and safe to perform and can give complete pain relief following abdominal incisions. If the incision is unilateral, it is sufficient to block on one side only; for midline incisions, however, the block must be bilateral. Usually 4-6 segments need to be blocked. Although the effect has been claimed to last as long as 24 hours, in our experience, the procedure has to be repeated every 6 hours and this makes it a labour intensive procedure requiring the presence of a practitioner able to administer the blocks.

Plexus blocks

i. Single shot

Plexus blocks, such as an axillary brachial plexus block, can give complete freedom from pain. Blocks administered immediately before or after surgery suffer the same problems as infiltration analgesia. In addition, even the use of dilute local anaesthetic solutions, sufficient to give analgesia, carry the risk of motor block. Trauma to an anaesthetic limb is also a risk.

ii. Continuous infusion block

A catheter inserted when establishing a plexus block can be used to maintain continuous analgesia of an area or limb. The problems of motor block and loss of sensation remain drawbacks.

Interpleural block

By inserting a fine catheter into the pleural space, an effective block of the intercostal nerves can be maintained. Analgesia and usually anaesthesia of one side of the abdomen gives good pain control following unilateral incisions. Interpleural block is not effective for midline or bilateral incisions, but bilateral blocks are generally not used. The risk of pneumothorax is small if a correct technique is used.

Epidural analgesia

Epidural catheters inserted at the level appropriate to the surgical incision can be used to provide a band of anaesthesia, with excellent pain relief. If the dose of the drug is adjusted correctly, it is often possible to avoid significant motor block or hypotension. Unfortunately, it is equally possible to produce either of these adverse effects, as well as urinary retention. The incidence of adverse effects makes the use of local anaesthetics alone for epidural analgesia unattractive. In practice, therefore, although epidural analgesia is used increasingly after major operations to achieve pain control, usually opioids are used. A small amount of local anaesthetic may be added as an adjuvant.

Management of postoperative pain in the drug dependent patient

Drug dependent patients present some problems in the management of postoperative pain. The postoperative period is usually not the appropriate time to undertake drug withdrawal. Patients who are narcotic dependent will exhibit tolerance to opioid analgesics and the usual doses will be insufficient to control their pain. If the patient is on a methadone program, their daily dose can be used to provide an index of their tolerance (the 'normal' analgesic dose of oral methadone is 10-20 mg). Otherwise, one can use repeated intravenous bolus doses of the chosen opioid until pain control is achieved, then adapt the dose to the chosen method of pain control. Good results with PCA in dependent patients have been obtained using often 10 mg or more as the bolus dose (instead of the usual 2 mg), with the normal 5 minute 'lock-out' and no 4 hour 'ceiling dose'. Certainly, some patients abuse this access to drugs, especially after the acute pain of the operation has started to subside, after 3, 4 or 5 days.

Epidural opioids are very useful in drug dependent and/or tolerant patients. One patient, on long-term morphine for carcinoma, required 210 mg of morphine, intravenously, administered over about 30 minutes until even moderate control of his postoperative pain was achieved. Finally, good control was achieved with 20 mg morphine, in 0.125% bupivacaine, 4 hourly by epidural injection. The usual dose for non-tolerant patients would be of the order of 8 mg daily.

Withdrawal from other substances of dependence, such as tobacco, alcohol or prescription drugs, also affects the management of postoperative pain. In one study, smokers required on average 30% more morphine in the postoperative period than non-smokers. Whether this was a withdrawal phenomenon or due to increased coughing, was not clear. Advice from specialists in the management of drug and alcohol problems is most useful in the management of these patients, especially at the time when acute postoperative pain begins to wane.

(See also Dental implications)

Further reading

Smith G. Postoperative pain. In: Nimmo WS, Smith G, editors. Anaesthesia. Oxford: Blackwell, 1989:1175-97.

Chrubasik J, Chrubasik S, Mather L. Postoperative epidural opioids. Berlin: Springer-Verlag, 1993.

Self-test questions

The following statements are either true or false.

1. Respiratory depression due to epidural morphine may develop more than 12 hours after injection.

2. The use of naloxone to reverse opioid-induced respiratory depression may cause severe pain.

Answers to self-test questions

1. True

2. True


  1. Austin KL, Stapleton JV, Mather LE. Multiple intramuscular injections: a major source of variability in analgesic response to meperidine. Pain 1980;8:47-62.
  2. Kehlet H. Stress free anaesthesia and surgery. Acta Anaesthesiol Scand 1979;23:503.
  3. Keeri-Szanto M, Heaman S. Postoperative demand analgesia. Surg Gynecol Obstet 1972;134:674.
  4. Woolf CJ, Chong MS. Preemptive analgesia - treating postoperative pain by preventing the establishment of central sensitization. Anesth Analg 1993;77:362-79.