Volume 167 Issue 1

Australian trends in opioid prescribing for chronic non-cancer pain, 1986-1996 James R Bell Readers may print a single copy for personal use. No further reproduction or distribution of the articles should proceed without the permission of the publisher. For permission, contact the Australasian Medical Publishing Company Journalists are welcome to write news stories based on what they read here, but should acknowledge their source as "an article published on the Internet by The Medical Journal of Australia <http://www.mja.com.au/>". Abstract - Introduction - Methods - Results - Discussion - Acknowledgements - References - Authors' details - ©MJA1997 Abstract Objective: To identify trends in the use of opioid drugs for chronic non-cancer pain. Design: Review of three sets of official records -- the record of Schedule 8 (S8) opioid drugs used in Australia, 1984-1995, from the national Department of Health, Housing and Community Services; New South Wales Department of Health statistical summaries of the number of authorities to prescribe S8 drugs for cancer pain and non-cancer pain for each June from 1990 to 1996; and NSW Department of Health patient records for a cohort of patients first prescribed S8 drugs in 1991. Main outcome measures: Total quantities of opioids used in Australia; numbers of S8 authorities issued in NSW. Outcome measures for the cohort study were the proportion of patients remaining on opioids long-term, the proportion for whom dose escalated over time, and the diagnoses for which opioids were being prescribed. Results: Between 1986 and 1995, the amount of oral morphine used in Australia rose from 117 to 578 kg. Use of all other oral S8 opioids combined increased from 93 to 149 kg. In NSW, the number of authorities to prescribe for non-cancer pain rose from 3326 in June 1990 to 5743 in June 1996 (73% increase), while cancer pain authorities rose from 2652 to 4831 (82% increase). Sixty-nine of the 102 patients ceased to receive drugs on authority over the five year follow-up. These subjects received opioids for a mean of 19 months. Among the remaining 33 subjects, dose escalation was common. Diagnostic information indicated that many patients had significant psychological and social problems. Conclusions: There has been a dramatic increase in opioid prescribing, a substantial proportion of which is for non-cancer pain. In a sample of patients being treated for non-cancer pain, long term use and dose escalation occurred in one third of cases. MJA 1997; 167: 26-29 Introduction There is evidence that the use of opioids is increasing worldwide,1 and the increasing use of slow release morphine in Queensland has recently been reported.2 In recent years there has been a cautious reappraisal of the role of opioids in chronic non-cancer pain.3,4 A recent survey of members of the Australian Pain Society found that 85% of respondents felt that opioid drugs could be used in patients with chronic pain.5 However, it is well recognised that benefits of opioid use in terms of improved pain control and sense of well-being need to be weighed against the risk of an increase in drug dependence. The aims of this study were to describe trends in opioid prescribing for chronic non-cancer pain, to investigate whether a person who begins taking oral opioids regularly will continue to do so in the long-term, and to assess the extent to which opioid treatment is associated with dose escalation over time. Methods Three data sources were used: Records from the Commonwealth Department of Health, Housing and Community Services. This organisation maintains records of drugs consumed in each State, and officers of the department provided data on national and State consumption of Schedule 8 (S8) opioid drugs for the years 1984-1995. Figures on oral opioid consumption obtained from the Commonwealth give a State by State breakdown on all morphine preparations except morphine mixtures. For the years 1986-1988 use of these morphine mixtures in each State was documented State by State, and NSW accounted for 50%-66% of national consumption; after 1988 only figures for national consumption are available. For the purposes of this study, NSW consumption of morphine mixtures after 1988 was estimated to be 55% of national consumption. By 1995 morphine mixtures accounted for less than 20% of total oral morphine consumption, so inaccuracy in this estimate is unlikely to alter greatly the overall trend in oral morphine use. Monthly summaries of authorities to prescribe S8 drugs from the Pharmaceutical Services Branch (PSB) of the NSW Department of Health. It is a legal requirement in all Australian States that regular prescribing of S8 drugs to any individual patient beyond a certain minimum period requires an authority. The issued authorities specify whether opioids are prescribed for cancer pain or non-cancer pain, and trends in the issuing of authorities provide an indication of whether there has been an increase in prescribing for non-cancer pain. In NSW, the PSB issues these authorities, and since 1990 the department has prepared a monthly summary of all current authorities. Summaries for the month of June each year were used to estimate trends in the point prevalence of authorities to prescribe opioid drugs from 1990 to 1996. Individual patient files held by the PSB. Applications for an authority to prescribe S8 drugs are reviewed by a medical committee, and new applications are usually approved if supported by someone with specialist medical qualifications. Most authorities are issued for 6 or 12 months, at which time a repeat application must be lodged, usually accompanied by a clinical report on the patient's progress. Thus, there is a record of reports and letters which have been submitted supporting applications to prescribe S8 drugs. From these records a cohort of patients was identified for whom a new authority (i.e., not a continuation of a previous authority) was issued in February and March, 1991, to prescribe an opioid drug in non-cancer pain. Records for this cohort were followed up for five years. Files were studied to ascertain in each case whether the patient continued taking S8 drugs for the full five years, whether their daily dose of opioid escalated over time, and to identify their diagnoses and comorbidities. The study was approved by the ethics committee of the South Eastern Sydney Area Health Service. The collection of data from NSW Department of Health files was performed by two medical students. No information identifying patients or prescribing doctors was available to any clinician who could have been involved in the patient's care. Results Opioid consumption The national consumption of opioids for 1986-1995 is displayed in the Figure. This illustrates a dramatic increase in the use of oral morphine, a similar rate of increase in the use of methadone syrup (employed in the treatment of addiction), and a small increase in other S8 opioids (primarily oxycodone and codeine). During the same period, use of injectable opioids also rose, from 4.1 to 7.1 million ampoules per year. The Figure (below) suggests that the advent of slow release morphine in 1990 in Australia has had a marked effect on prescribing patterns, as the rate of increase in morphine prescribing accelerated from that point. However, although some individual opioid drugs have declined in use, the increase in oral morphine has not been accompanied by a decline in the use of other opioid drugs. Figure: Australian consumption of opioids, 1986-1995. Figures for NSW parallel the Australian rise. Between the years 1990 and 1995, total oral morphine use in NSW rose from 87 to 209 kg per year (a 140% increase). In the same interval, the consumption of other S8 drugs rose from 42 to 46 kg per year (10% increase). Injectable S8 drugs rose from 1.2 to 1.3 million ampoules (8% increase). Authorities The number of S8 authorities current each June in NSW rose between 1990 and 1996. In those years, the number of authorities to prescribe for non-malignant pain rose from 3326 to 5743 (73% increase), while cancer pain authorities rose from 2652 to 4831 (82% increase). Follow-up of authority applications There were 102 patients for whom new authorities to prescribe an S8 opioid for non-malignant pain were sought in February and March, 1991: 54 women (mean age, 57; range, 28-91) and 48 men (mean age, 51; range, 23-77) (see Box). Although these were new applications, 30 subjects were already receiving other S8 opioid drugs on authority. The clinical impression is that dose escalation may be more likely in patients taking the potent, short-acting drug dextromoramide, and less likely in those taking low potency codeine or long-acting methadone. Despite the small numbers (see Box), the difference in the proportion of patients taking these different drug types whose opioid dose escalated approached significance (P = 0.053). Thirty-three patients (32.4%; 95% CI, 23.4%-42.3%) were still receiving S8 opioids five years later (although many were receiving a different S8 opioid in 1996). Twenty-six of the 102 patients (25.5%; 95% CI, 17.4%-35.1%) received escalating doses of opioids: 21 of the 33 patients who were still receiving opioids on authority in 1996, and five of the 69 people not receiving opioids on authority in 1996 (chi-squared = 44.69; 1df; P < 0.005). Patients receiving opioids for the full five years were more likely to escalate their dose (odds ratio, 22.4; 95% CI, 7.1-71.0). It is unlikely that escalation was simply a function of patients beginning with a low initial dose of opioid. Converting initial doses to equivalents of oral morphine, the group whose doses did not escalate began with a mean dose of 80 mg of oral morphine per day, while the group whose doses did escalate began with a mean dose of 87 mg per day. The PSB records documented that 14 patients from this cohort died between 1991 and 1996 (7 men, 7 women; mean age, 70; age range, 42-91). Most were elderly, with serious underlying disease. It seems that in many of these patients, prescribing opioids was part of terminal care. Excluding subjects known to have died, 33 of the remaining 88 subjects (38%) were still receiving opioid drugs on authority five years later. The 55 no longer receiving opioids on authority had received opioids for a mean of 19 months. Clinical data from authority applications In 63 applications the primary diagnosis was musculoskeletal pain, specifically back pain in 48 of these. In five cases the diagnosis was simply "chronic pain". Several patients had multiple, unrelated diagnoses -- such as "chronic back pain and gunshot wound", "backache and migraine", or "myalgic encephalomyelitis and migraine". In 27 cases seemingly unrelated diagnoses were recorded. The presence of multiple sources of severe pain probably reflects the prevalence in this population of somatoform disorders (i.e., disorders in which physical complaints are not fully explained by known medical conditions, or in which complaints or impairment are in excess of what would be expected from the patient's medical condition). The comments in the patients' files suggest that in many cases prescribing opioids is a response to difficult and unmanageable problems. There is more than a hint of frustration in many of the comments -- such as the terse letter: The situation is unchanged. His back feels "like a bag of broken china". I am still at a loss to know what to do and I have arranged to see him again in six weeks with further x-ray. Or a general practitioner's bleak summation: This patient has been so adversely affected by her CFS [chronic fatigue syndrome] and migraine she has been rendered completely dysfunctional as a human being and mother. While many of the patients in this study had multiple medical problems and histories of extensive treatment, often there were strong intimations of concurrent psychological and adjustment difficulties. For example, one patient had a history of multiple back operations, with a background of social problems described as "marriage collapsing, alcoholic father living with them, and a history of overdoses on tricyclic antidepressants and on benzodiazepines". While some of the patients had clearly defined diseases, and several were seriously ill, in 62 cases (61%) there were poorly defined medical problems. In 22 cases there was no clinical information other than a diagnosis. Among the other 80 files, the presence of social and emotional problems was noted in 29 cases (36%). The quality and thoroughness of medical information documented in the health department files was very limited. Discussion There has been a dramatic increase in the prescribing of oral morphine over the past decade. To put this rise in perspective, it is useful to compare the increase in prescribed morphine with the increase as a result of government policy in prescribed methadone syrup. Between 1986 and 1995 there was a marked expansion of the methadone program for the treatment of addiction, with the number of patients in treatment Australia-wide rising from around 1000 in 1984 to 13 000 in 1995. However, the increase in morphine consumption, driven by changes in clinical practice rather than policy, has been comparable to the increase in consumption of methadone syrup. Total use of other S8 opioids also increased during this period, suggesting that morphine consumption has not grown through replacing another opioid. The national data on overall opioid use are an accurate measure of the quantity of these drugs being prescribed each year. However, there are limitations on the use of official records in determining how much of this increase was for terminal care, and how much for chronic pain. In this study, the increase in authorities current in NSW was about half as great as the increase in oral morphine use between 1990 and 1995. Given that oral morphine is seldom used in acute pain management, the disparity between authorities and total amount of drug used suggests that many patients receive morphine for prolonged periods without an authority to prescribe having been issued. There is no way of knowing how many doctors prescribe opioids long-term to individual patients without ever seeking authority. Furthermore, it is possible that there is a systematic bias, with doctors prescribing for terminal care being less likely to apply for an authority than when prescribing for chronic non-cancer pain. However, even with these reservations, it seems reasonable to conclude that the increase in authorities in both categories and the increase in overall use of morphine indicate a change in attitudes to opioids, with more liberal use in both terminal care and chronic pain. Might this more liberal approach contribute to a problem with iatrogenic drug dependence? The hallmark of opioid dependence is that it is a chronic, relapsing, long-term problem. Most of the 102 patients who began taking opioid drugs on authority in early 1991 were no longer doing so five years later. However, some of them may have continued to receive opioids without an authority. Patients who move interstate, find a new doctor who does not seek an authority to prescribe or who "doctor shop" to get prescriptions from multiple prescribers may not appear in the health department records even though they continue to receive opioids. Some may have switched to agents such as combined paracetamol-codeine, a drug containing a modest dose of opioid, but for which no authority is required. It is also possible that some patients died without this coming to the attention of the health department. However, even allowing for these possibilities, the figures suggest that fewer than half the surviving patients were still taking opioids regularly five years after first taking them. It is difficult to assess the extent of problems of tolerance and dependence from the incomplete clinical data in official records. Previous studies on selected groups being treated in specialist facilities have reported that opioids can be used safely and effectively,3,4 but these reports give little indication of the extent to which problems are likely to arise in more diverse samples of patients treated in a range of settings. Dose escalation may not be a reliable indication of dependence, but may reflect changes in pathology. However, previous reports indicate that most patients with non-cancer pain do not require escalating doses over time.6 Clinical experience with cancer pain management indicates that during the first weeks of dosing with opioids a degree of tolerance is reached, but this reaches a plateau and thereafter a stable dose can be beneficial for prolonged periods.7 The system of S8 authorities means that patients have usually been receiving opioids for a couple of months before authority application, and by then should usually be receiving a stable and effective dose. Most patients discontinued opioid use after quite prolonged periods without dose escalation. It is likely that among the 25% whose dose escalated there were many patients with problems of dependence. The comments and diagnoses in the files indicate that opioids are often prescribed for patients with social problems, high levels of emotional distress, and unclear medical diagnoses. It is likely that many of these subjects have a mixture of somatoform disorders and diverse medical and psychosocial problems. The benefits and risks of opioid use in this setting are poorly understood. A supportive doctor-patient relationship and attention to psychiatric and medical comorbidity are more important aspects of the management of somatoform disorders.8 The value of psychological treatment has been demonstrated in recent randomised trials.9 The implication for treating chronic pain states where there appear to be contributing psychological or social factors is the need for a comprehensive approach to treatment. Opioid drugs may have a useful role in such a comprehensive approach. The risks of opioid dependence, with escalating pain and drug use and diminished activity, need to be set against the potential benefits in terms of improved pain control and well being. It has been suggested that the risks of dependence can be minimised when the drugs are prescribed by experienced and skilled medical practitioners.3 It is also probable that the current system of requiring doctors to obtain authorities to prescribe opioids long term is a valuable restraint on inappropriate prescribing of these drugs. Acknowledgements Data from the NSW Department of Health files were gathered by two medical students (Amanda Mather and Sarah Baldwin). Mr John Lumby, Pharmaceutical Services Branch, NSW Department of Health, assisted with access to Branch files. Ms Cheryl Regan, Treaties and Monitoring Section, Commonwealth Department of Health and Family Services, assisted by providing summary data on opioid consumption in Australia. The federal and State departments do not necessarily endorse any of the conclusions or opinions drawn from any data. References Hiraga K, Mizuguchi T, Takeda F. The incidence of cancer pain and the improvement in pain management in Japan. Postgrad Med J 1991; 67 Suppl 2: S14-S25. Richards AH. The use of controlled-release morphine sulfate (MS Contin) in Queensland 1990-1993. Med J Aust 1995; 163: 181-182. Portenoy RK, Foley KM. Chronic use of opioid analgesics in non-malignant pain; report of 38 cases. Pain 1986; 25: 171-186. Zenz M. Morphine myths; sedation, tolerance, addiction. Postgrad Med J 1991; 266: 2392-2397. Australian Pain Society. Newsletter, July 1996. McQuay HJ. Opioids in chronic pain. Br J Anaesth 1989; 63: 213-226. Foley KM. Changing concepts of tolerance to opioids: what the cancer pain patient has taught us. In: Chapman CR, Foley KM, editors. Current and emerging issues in cancer pain; research and practice. New York: Raven Press, 1993: 331-350. Quality Assurance Project. Treatment outlines for the management of the somatoform disorders. Aust N Z J Psychiatr 1985; 19: 397-407. Guthrie E. Psychotherapy of somatization disorders. Curr Opin Psychiatr 1996; 9: 182-187. (Received 11 Oct 1996, accepted 6 Mar 1997) Authors' details Drugs and Alcohol Unit, Prince of Wales Hospital, Sydney, NSW. James R Bell, BA, FRACP, Director. No reprints will be available from the author. Correspondence: Dr J R Bell, Drugs and Alcohol Unit, Prince of Wales Hospital, High Street, Randwick, NSW 2031. E-mail: james.bell @ unsw.edu.au ©MJA 1997 <URL: http://www.mja.com.au/> © 1997 Medical Journal of Australia.

James R Bell

The use of oral opioids in patients with chronic non-cancer pain Management strategies Paul J Graziotti and C Roger Goucke Readers may print a single copy for personal use. No further reproduction or distribution of the articles should proceed without the permission of the publisher. For permission, contact the Australasian Medical Publishing Company Journalists are welcome to write news stories based on what they read here, but should acknowledge their source as "an article published on the Internet by The Medical Journal of Australia <http://www.mja.com.au/>". Abstract - Introduction - Which patients? - Which doctors should prescribe? - Which drug? - Consent - How should these drugs be prescribed? - References - Authors' details - ©MJA1997 Abstract Background: The use of oral opioids in non-cancer pain is increasing, but it is not clear that this is improving outcomes for patients. These management strategies were developed as a consensus view between the two authors, who are both Directors of the Australian Pain Society. The strategies were subsequently reviewed and approved by the other Directors of the Society: four anaesthetists specialising in pain management, a pharmacist, a rheumatologist, two rehabilitation physicians and an occupational therapist. Evidence base: A Medline search of the literature since 1966 produced 163 relevant articles, including two randomised controlled trials of oral opioids in non-cancer pain. Management consensus A small group of patients with chronic non-cancer pain can benefit from the use of oral opioids. Thorough attention to diagnosis and patient history must precede any decision to prescribe opioids. Patients should be psychologically stable. Patient and doctor should agree beforehand on how to assess the outcome of therapy. Only one doctor (the patient's regular primary carer or pain specialist) should prescribe opioids and assess the response. Sustained release morphine preparations are the drug of choice. A trial of therapy, with goals and endpoint agreed between patient and doctor, should precede any decision to prescribe opioids in the long term. MJA 1997; 167: 30-34 Introduction Patients with chronic non-cancer pain present a number of challenges to their treating physicians. One such challenge is whether to use oral opioids in their treatment plan. The combination of poorly defined pathology, significant psychosocial factors, manipulative behaviour, dependence, tolerance and government regulations are formidable influences on management decisions. Because of these concerns, many doctors may decide not to prescribe opioids for patients with chronic non-cancer pain, and certainly current legislation discourages it. Arguments against the use of opioids in these patients have also been published.1,2 Yet, world-wide, there is a growing body of opinion that a small subgroup of patients with chronic non-cancer pain may function better and have less pain if treated with opioids, without requiring rapidly escalating doses or showing addictive behaviour. Data from recent randomised controlled trials3-5 support the finding of benefit in retrospective studies of patients treated with opioids for chronic non-cancer pain.6,7 This evidence suggests that a proportion of patients report an improvement in their level of analgesia and/or level of function.8 The prevalence of drug abuse, dependence and addiction has been estimated to be as low as 3.2%9 or as high as 18.9%,10 depending on definition. Fishbain et al. report that there is little evidence that addictive behaviours are common in the chronic pain population.11 There is evidence, however, that an increasing number of Australian patients are receiving prescribed oral opioids for both malignant and non-malignant pain.12 This may be filling a previously unmet need, but it is not clear if the increased prescribing for non-cancer pain is appropriate, whether there has been any increase in function, reduction of pain, reduction in suffering, or if any of these drugs will enter the illicit market. How then can we ensure the maximum benefit from the prescription of opioids for chronic non-cancer pain? Who are the appropriate patients? Who are the appropriate prescribing doctors? How should the drugs be prescribed? Which drugs should be prescribed? How and when should they be withdrawn? The aim of this article is to explore these issues and provide guidelines to assist practitioners in the appropriate use of oral opioids. Which patients? It is essential that all reasonable attempts be made to achieve a diagnosis for the cause of the pain, including nociceptive, neuropathic and psychological contributions. The demonstration of pathology commensurate with the degree of pain behaviour is desirable. However, patients often have pathology which is difficult to interpret (e.g., degenerative changes on spinal x-rays). Certain conditions result in neuropathic pain, which is usually a clinical diagnosis and may not be reflected in investigations such as radiographs or nerve conduction studies. This should not preclude a trial of opioids in these patients if otherwise appropriate. A thorough history of previous conservative therapy ( Box 1) should be taken before consideration is given to the medium to long term use of opioids. Previous drug therapy should include trials of non-opioid analgesics, tricyclic antidepressants and membrane-stabilising medications (e.g., sodium valproate, carbamazepine, mexiletine). Patients for whom opioids are being considered should be psychologically stable, although it is recognised that this is difficult to define. Patients in chronic pain may develop psychological problems as a result of the pain, and therein lies a dilemma for the physician. Will treating the pain reverse some of the psychological abnormalities? Or are the psychological abnormalities a significant contributor to the overall pain behaviour? Studies would suggest the former in most cases.13 A psychological assessment is essential for patients with poorly defined pathology, younger age, high levels of distress, or previous or ongoing substance abuse. Consideration should be given to managing these patients in a multidisciplinary pain centre ( Box 2). Which doctors should prescribe? The prescribing doctor should have an established therapeutic relationship with the patient. This excludes casualty officers, specialists who see a patient on one occasion only, after-hours locum services and (usually) junior medical staff working in outpatient departments. It is important that only one doctor prescribes the opioids and assesses the response. Patient and doctor must agree on how to assess the outcome of therapy before opioids are prescribed. Failure to reach the predetermined goals is an indication to cease prescribing. All patients who are considered suitable for the long term use of opioids in non-cancer pain should be assessed at some stage in a specialist pain management centre. Shared care between the general practitioner and the pain management centre is ideal. Which drug? Sustained release morphine preparations are the drugs of choice in patients with chronic non-cancer pain, because of their single- or twice-daily dosage and stable blood con centrations as a consequence of their more predictable pharmacokinetics. There is agreement internationally14 and within Australia15 that intramuscular opioids should not be used to treat chronic non-cancer pain. In particular, intramuscular pethidine should be avoided. It has a short half-life, possibly an increased risk of dependence due to its psychomimetic effects, and the potential for excitatory central nervous system effects from accumulated norpethidine concentrations after repeated doses. Codeine phosphate is a short-acting drug and as such has little place in chronic pain management. A controlled release preparation is available overseas and may prove useful. Immediate release morphine as morphine mixture (5-10 mg/mL) may be used for dose finding before beginning the use of sustained release morphine, but is generally unnecessary. It may be useful for breakthrough pain or exacerbations. Transdermal fentanyl patches may, in the future, provide a useful alternative for patients intolerant of morphine. Methadone and oxycodone rectal suppositories are useful alternatives to oral sustained release morphine preparations. Dietary advice should be given to minimise the problem of constipation and consideration should be given to the regular use of laxatives. Consent Patients prescribed opioids for the treatment of chronic non-cancer pain should be fully informed of the potential consequences of this therapy. There is increasing awareness that a written consent form* is a valuable tool, particularly when treating patients who for any reason are difficult to manage.16 Informed consent should include discussion of: Clearly defined specific goals of the treatment program. The likelihood of dependence and the risk of addictive behaviour. All patients will become dependent and are likely to experience withdrawal symptoms if opioid therapy is suddenly stopped. Addictive behaviour occurs in a much smaller proportion of patients and may be minimised by appropriate patient selection. The lack of published data on long term outcome of the effects of medically prescribed opioids. The potential for cognitive impairment, in particular that might affect driving ability. While there are few studies in non-cancer patients assessing cognitive function in the presence of opioids,17 the studies in cancer patients would suggest that cognitive function is actually improved when adequate analgesia is provided.18 The potentiating effect of opioids on the sedative effect of other medication. The possibility (for women) of physical dependence in children born to them if they continue to take opioids in late pregnancy. Indications for the cessation of treatment with opioids. The patient's responsibilities regarding the security of his or her medication. The consequences of aberrant behaviour ( Box 3) should be identified as clearly as possible. Side effects (e.g., constipation, nausea, sedation, dry mouth). How should these drugs be prescribed? Trial of oral opioid Before prescribing opioids on a long term basis, a trial should be undertaken over four to six weeks. Goals for the trial should be identified between doctor and patient, and the endpoint clearly stated. One doctor should institute and monitor the trial. The trial should commence with the equivalent of sustained release morphine 10-50 mg twice a day, with the outcome assessed after one week or less. Depending on response, the dose may be increased or decreased. In general, round-the-clock medication is the accepted regimen, although in patients with fluctuating pain conditions it may be more appropriate to consider a variable dosing regimen with oxycodone or morphine elixir.19 Patients must accept the responsibility of ensuring their supply of medication does not run out after hours. There is controversy regarding the expectation that patients will improve in function. Is it adequate for patients to achieve analgesia only? Is it adequate for patients to state that they feel better only? Certainly patients should reduce their use of other analgesics; ideally, they should show improved function. Perception of improved analgesia should be the minimal requirement, and failure to achieve at least partial analgesia at a moderate dose contraindicates any further long term opioid treatment. Most patients who experience minimal or no analgesic effect will stop taking the drug themselves before the end of the trial. Similarly, many patients who experience adverse side effects, such as severe nausea or constipation, will determine that these outweigh the analgesic benefits and stop taking the drug. Opioid-naive patients whose dose rapidly escalates within a month of starting treatment should generally be considered inappropriate for long term opioid therapy. At the end of the trial period, if the expected outcomes have not been achieved, the drug dose should be tapered over a few days and ceased. Ongoing reviews Patients who are then prescribed opioids on an ongoing basis should be reviewed at first fairly frequently (e.g., weekly), then monthly, by the prescribing doctor. A detailed review by a pain management centre should be undertaken annually. At each review, analgesic efficacy should be assessed, as should any improvement in the level of function. The responsible federal or State health department must be notified. Evidence of aberrant behaviour should be assessed. Aberrant behaviour is variable in its importance and relevance. Box 3 indicates factors which Portenoy20 has considered more or less predictive of the development of addictive behaviour. Patients identified with behaviour in the less predictive category indicate a need to assess the dose of drug, the psychological factors of relevance, the patient's expectations or the type of medication. If patients are identified with features in the more predictive category, the appropriateness of opioid prescription should be seriously reassessed. Often it will be necessary to reduce and then withdraw the opioid over a week. In other cases, a more regulated supply, such as daily or weekly prescriptions, may be appropriate. An initial written consent form indicating those factors for which supply will be withdrawn will make this easier. References Large RG, Schugg SA. Opioids for chronic pain of non malignant origin: caring or crippling. Health Care Anal 1995; 3: 5-11. Butler SH. Opiates for chronic pain: present American controversy. Regul Pept 1994; 52 Suppl 1: S295-S296. Jaddad AR, Carroll D, Glynn CJ, et al. Morphine responsiveness of chronic pain: double blind randomised crossover study with patient controlled analgesia. Lancet 1992; 339: 1367-1371. Arkinstall W, Sandler A, Goughnour B, et al. The efficacy of controlled release codeine in chronic non-malignant pain: a randomised placebo controlled trial. Pain 1995; 62: 169-178. Moulin DE, Iezzi A, Amireh R, et al. Randomised trial of oral morphine for chronic non-cancer pain. Lancet 1996; 347: 143-147. Portenoy RK, Foley KM. Chronic use of opioid analgesics in non-malignant pain: report of 8 cases. Pain 1986; 25: 171-186. Zenz M, Strumpf M, Tryba M. Long term oral opioid therapy in patients with chronic non-malignant pain. J Pain Symptom Manage 1992; 7: 69-77. Jamison RN. Comprehensive pretreatment and outcome assessment for chronic opioid therapy in non-cancer pain. J Pain Symptom Manage 1996; 11: 231-241. Tanb A. Opioid analgesics in the treatment of chronic intractable pain of non- neoplastic origin. In: Kitahata LM, Collins D, editors. Narcotic analgesics in anethesiology. Baltimore. Williams & Wilkins, 1982: 199-208. Tennant FS, Robinson D, Sagherian A, Seecof R. Chronic opioid treatment of intractable, non-malignant pain. NIDA Res Monagr 1988; 81: 174-180. Fishbain DA, Rosomoff HL, Rosomoff RS. Drug abuse, dependence and addiction in chronic pain patients. Clin J Pain 1992; 8: 77-85. Richards AH. The use of controlled-release morphine sulfate (MS Contin) in Queensland 1990-1993. Med J Aust 1995; 163: 181-182. Waddell G, Pilowski I, Bond MR. Clinical assessment and interpretation of abnormal illness behaviour in low back pain. Pain 1989; 39: 39-41. Hagen N, Flynne P, Hays H, McDonald N. Guidelines for managing chronic non-malignant pain. Canadian Fam Physician 1995; 41: 49-53. Cherry DA, Gourlay GK. Pharmacological management of chronic pain: A clinician's perspective. Agents Actions 1994; 42: 173-174. Burchman SL, Pagel PS. Implementation of a formal treatment agreement for outpatient management of chronic non-cancer pain with opioid analgesics. J Pain Symptom Manage 1995; 10: 556-563. Zacny JP. A review of the effects of opioids on psychomotor and cognitive function in humans. Exper Chem Psychopharm 1995; 3: 432-466. Vainio A, Ollila J, Matikainen E, at al. Driving ability in cancer patients receiving long term morphine analgesia. Lancet 1995; 346: 667-670. Savage SR. Long term opioid therapy: assessment of consequences and risks. J Pain Symptom Manage 1996; 11: 274-286. Portenoy RK. Opioid therapy for chronic nonmalignant pain: a review of the critical issues. J Pain Symptom Manage 1996; 11: 203-217. Authors' details Department of Pain Management, Sir Charles Gairdner Hospital, Perth, WA. Paul J Graziotti, FANZCA, FFARCS, Visiting Specialist; West Australian Director, Australian Pain Society. C Roger Goucke, FANZCA, Head; Secretary, Australian Pain Society. No reprints will be available from the author. Correspondence: Dr C R Goucke, Department of Pain Management, Sir Charles Gairdner Hospital, Perth, WA 6009. ©MJA 1997 <URL: http://www.mja.com.au/> © 1997 Medical Journal of Australia.

Paul J Graziotti