Drug and Alcohol Findings home page in a new window Background notes

Randomized controlled trial of dexamphetamine maintenance for the treatment of methamphetamine dependence


Featured study's methodology

It seems likely that the study attracted an atypically motivated and 'together' set of patients. It started with 270 prospective patients whose methamphetamine use might have warranted inclusion in the trial. That just 49 were actually included is indicative partly of the extra demands Such as the need to have testable hair, accept random allocation, attend for screening, and be prepared to be withdrawn after 12 weeks on the medication. of the research process, but perhaps also of the limited attractiveness and applicability of a daily supervised maintenance regimen for stimulant-dependent patients. Their need to avoid withdrawal and maintain continuity of consumption is not as great as those dependent on opiate-type drugs, typically they (especially injectors) value the initial pleasurable 'hit' which tablets do not give them, and frequent stimulant use is often associated with a disordered lifestyle, antithetical to daily attendance for supervised consumption.

Though the authors attribute the dramatic improvements in both groups to the support and structure offered by the study's treatments, it seems equally likely that people willing to give this structure a try and to accept the extra burdens of the research were already among the most promising and motivated of patients. The fact that they did so well, even when many did not take advantage of the offer of psychotherapy, seems to support this reasoning. A looser prescribing regimen without supervised consumption or research add-ons might have attracted less motivated patients who would not have done so well.

That improved retention and other advantages of amphetamine over placebo were indeed conferred by the medication is indicated not just by the data, but by the plausibility of this as an explanation for the results. However, there are alternative explanations and other possible contributors to the findings. The gap in drop-out rates emerged in the first two to three weeks of the trial as the stabilisation period was ending and treatment proper was beginning, presumably as patients began to feel the effects of differences between the two sets of tablets. Though barely more than half the placebo group guessed their medication correctly, 7 in 10 of the those on methamphetamine did so. This breach in the 'blinding' of patients may have resulted in more widespread demoralisation and disappointment in the placebo group, giving methamphetamine an advantage which it would not necessarily have in normal clinical practice.

Even patients who could not be reassessed were included in the analyses of outcomes, an important way to maintain comparability between the two groups. The article does not specify what assumptions were made about how these patients would have fared had they been able to be reassessed. Had it had been assumed that their dependence remained the same as when last measured, the gap in the re-assessment rate of 17% in favour of amphetamine would probably have contributed to the medication's advantage compared to placebo. By chance, on several measures the placebo group could be expected to have had a worse prognosis. On average they had been using methamphetamine a year longer, were more likely to have previously been treated for this use, 14% extra injected the drug – a route the study found associated with worse outcomes – and they were less likely to be employed. None of these differences were statistically significant, but cumulatively they might have tipped the balance slightly in favour of amphetamine.

The significantly greater easing of dependence at the two-month follow-up seemed an important persisting effect of amphetamine prescribing, but this comparison was not planned in advance of the results of the study being known. Post-hoc analyses Post-hoc analyses of this kind are best seen as generating hypotheses for testing in a study specially designed for this purpose. The main problems are that researchers can select comparisons to highlight on the basis that these were the ones that produced statistically significant differences, greatly increasing the risk these differences were due to chance. of this kind can only be considered suggestive until confirmed in a study intended for this purpose.

Amphetamine users in treatment in Britain

Across the UK in 2009/10 or 2010, at least 12,128 people using amphetamine were in or referred for treatment for problems with illegal drugs, just over 5% of the total of 226,302. These totals are an amalgam of differently defined statistics in the different nations. In England in 2009/10, 3858 of the nearly 207,000 adults treated for illegal drug use had amphetamines recorded as their main problem drug and another 7259 as a secondary drug. In Wales in the same year, 732 of 11,216 adults referred for treatment for illegal drug use problems had a primary amphetamine problem. In Scotland, amphetamine was recorded in the drug use profiles of 279 of the 8086 patients starting treatment for problems involving illegal drugs.

British studies

1 The only UK trial to rigorously test amphetamine prescribing by randomly allocating amphetamine-dependent patients to this plus a psychosocial treatment package, or to the same package but without prescribing, took place in Manchester and Cardiff and was published in 2004. The aim was to test whether prescribing added anything to a "best available" treatment package. Consisting of provision of literature on amphetamine, motivational style interviewing, review of recent behaviour using a retrospective drug diary and discussion of cues, coping and lapse management, advice on healthy lifestyles, harm minimisation advice including advice on safer injecting and use of syringe exchange schemes, referral to appropriate non-drug agencies for other health or social issues, symptomatic prescribing for depression, anxiety and insomnia and the possibility of inpatient admission for detoxification if clinically indicated. Up to 100mg dexamphetamine tablets a day were dispensed daily through community pharmacies for the first four months of the seven-month trial, then were gradually withdrawn. Researchers independent of the treatment team assessed the 59 patients (of whom 56% were injecting) in the trial at periods up to two months after treatment ended.

There was no substantial or statistically significant benefit associated with prescribing. In particular, both groups reduced their illegal amphetamine use and their rates of injecting to roughly equivalent degrees. There was, however, some evidence of a greater curtailing in the range of drugs used by each individual (an interpretation of the term "polydrug" use) and of greater improvements in physical and psychological health. Prescribing dexamphetamine did not have adverse effects on physical or psychological health. There was only one episode of psychosis, this experienced during the dexamphetamine reduction phase by a prescribed patient under severe stress.

This study suffered from a high drop-out rate; just 56% of patients could be reassessed at the end of treatment and 59% at the final follow-up. It is unclear from the brief report available how representative the patients were, but it seems likely that they were highly self-selected and that the trial also imposed exclusion criteria. These two factors alone mean the results may not be applicable to the general run of potential patients. Also, prescribing was added to what was and is probably an unusually good treatment package; the tablets may have made a greater difference if psychosocial support have been weaker. Four months would normally be considered insufficient for a maintenance prescribing programme to create and embed substantial lifestyle changes. Like the other randomised trial in Australia, the UK trial could as justifiably be considered stabilisation prior to withdrawal as maintenance. For these reasons the trial cannot be considered a definitive verdict on the (non-)value of amphetamine maintenance.

For the authors the message was that high quality and comprehensive psychosocial treatment could attract and benefit amphetamine users, while dexamphetamine substitution should remain a specialist intervention carried out only by experienced practitioners as part of a complete treatment package. This should incorporate psychosocial interventions and clinical monitoring procedures, including urine drug screening with the ability to differentiate prescribed from illicit amphetamines, blood pressure checks, and mental state reviews.

2 A report from the Midlands advocated for amphetamine prescribing on the basis of experience with 32 patients prescribed the drug from 1993–2005. Given concerns over precipitating a psychotic episode, a key finding was that for the five patients who started treatment with psychotic symptoms, the opposite was the case; they all improved after starting amphetamine treatment and stabilising their doses. Receiving pure dexamphetamine in low controlled doses contributed, the authors, thought to their remission. After on average nearly two years in treatment, a quarter had left after being withdrawn and nearly half had dropped out.

3 From a report published in 2002, findings were similar in Manchester, where eight schizophrenic, non-violent high-dose amphetamine users were prescribed dexamphetamine after they had proved unable to stop using the drug. Prescribing was carefully titrated up from a conservative starting point and was contingent on cooperation with psychiatric treatment for their mental illness. In none was their mental illness exacerbated by the prescribing, and six experienced at least partial mental health and substance use benefits. For this refractory and difficult set of patients, "The huge advantage of the treatment ... was that [they] were seen regularly and, for the most part, took their neuroleptic medication regularly. As a result, their mental state improved, and they were able to receive the benefits of advice and support directed at health and welfare needs".

4 In contrast, mental health problems remained common among patients in Wolverhampton. The records of 20 patients treated with dexamphetamine from 1984 to 1998 (10 were still being treated at the end of this period) revealed that a quarter had developed paranoid or psychotic symptoms while on the prescription, generally due to using illicit amphetamine or stockpiling and then 'bingeing' on their prescribed doses. Symptoms remitted when illicit injecting stopped. Though a "serious concern", the clinicians who authored the study were reassured that the prescription itself did not cause psychotic episodes. How many of these patients might have developed psychotic episodes even in the absence of a prescription is unclear. What does seem clear is that those retained in treatment by the prescription were in a better position to have their relapse identified and treated, and that injecting and illicit drug use often fell while patients were in treatment. The authors concluded that "The present study provides some support for the experience of many practitioners that it may be possible to retain patients in treatment for prolonged periods and thereby reduce psychosocial and physical harm".

5 In the early '90s, a South Wales community drug team compared the progress of 63 clients prescribed dexamphetamine against that of 25 who also fulfilled the study's criteria (primarily that they injected amphetamine) but had attended the service before dexamphetamine prescribing began. This attempt to engineer a comparable control A group of people, households, organisations, communities or other units who do not participate in the intervention(s) being evaluated. Instead, they receive no intervention or none relevant to the outcomes being assessed, carry on as usual, or receive an alternative intervention (for the latter the term comparison group may be preferable). Outcome measures taken from the controls form the benchmark against which changes in the intervention group(s) are compared to determine whether the intervention had an impact and whether this is statistically significant. Comparability between control and intervention groups is essential. Normally this is best achieved by randomly allocating research participants to the different groups. Alternatives include sequentially selecting participants for one then the other group(s), or deliberately selecting similar set of participants for each group. group was only partially successful, and the groups differed on several variables.

The first finding was that the availability of prescribing may have attracted more amphetamine injectors in to treatment. Comparing the 29 months before prescribing to roughly the same period after, numbers more than doubled and the proportion they formed of the service's caseload also nearly doubled from 24% to 43%. Retention also dramatically improved from an average of one and a half months to nearly a year (and probably longer because 23 patients were still in treatment), and the drop-out rate plummeted from 80% to 37%.

So rapidly did pre-prescribing clients drop out that it was not possible to reassess them to measure behaviour change. Among post-prescribing patients, illegal amphetamine use was cut to less than a third of its pre-treatment level and injecting and infection risk behaviour also fell substantially. Three of the 63 patients experienced psychotic episodes related to heavy stimulant use 'on top' of prescribed amphetamine, but none were attributable to starting treatment nor to the prescribed drug alone. The per month incidence of psychosis was much less than in the comparison group. There were several non-serious cases of raised blood pressure, the worst again related to supplementary substance use.

The authors admitted that "By the time the evaluation ... all those involved with the programme were persuaded of its worth for their clients on the evidence of experience". Since they collected the data they may have been biased, but it seems likely that their enthusiasm was due to real benefits for their patients which were also reflected in the study. The conclusion was that for this selected group of high-risk amphetamine users, prescribing brought substantial benefits at the cost of no major adverse side effects.

6 In Cornwall in England's south west region, researchers examined the records of 148 primary amphetamine users (of whom 95 injected) prescribed amphetamine elixir by the community drug team during 1992–1996 – the largest sample studied in Britain. They were a subset of the 220 clients prescribed the drug from the roughly 500 amphetamine-linked referrals to the team over four years. Medication was dispensed at local pharmacies in arrangements which varied from daily supervised consumption to weekly pick-up and doses (at their peak for each individual) averaging around 50mg. Typically clients stayed in treatment for five (injectors) or seven (oral users) months. The authors were concerned at the high throw/drop-out rate (35% of injectors were discharged for non-attendance and 52% dropped out), but commented that in the absence of substitute prescribing, early drop-out was virtually universal.

While in treatment, around 60% of patients achieved at least a couple of weeks' abstinence from amphetamines and all other drugs except cannabis and alcohol, and over 80% 14% of injectors and 9% of oral users relapsed into street-use after successfully stopping. seemed to have sustained this. Of the injectors, over 6 in 10 stopped injecting, usually within two months of starting treatment.

On the crucial issue of mental health, of the 220 clients prescribed amphetamine, none without a history of psychosis experienced a psychotic episode. Of the seven with a history, four diagnosed schizophrenic proved unable to adhere to the strict treatment contract imposed by the service and were discharged. Two of the other three stopped using street drugs and were subsequently symptom-free. The third continued to inject street drugs and experienced a psychotic relapse. The authors concluded that the dangers of amphetamine prescribing have probably been over-estimated. Out of a treatment sample of 220 being monitored for the equivalent of 2640 person-months and prescribed at higher levels than is the norm in other services, only five psychotic episodes were recorded, all in people who had a previous history and had continued to inject street drugs; none appeared related to prescribing alone.

7 Also in Cornwall and at the same community drug team, the progress of 60 amphetamine injectors prescribed dexamphetamine was compared with that of 120 heroin injectors prescribed methadone. All had been discharged during 1995 and 1996. Apart from the drug being prescribed, other treatment elements including the dosing strategy and psychosocial support were similar. Dexamphetamine averaging 43mg a day was dispensed from community pharmacies. Clinic appointments were fortnightly.

The outcomes were in some respects remarkably similar. On dexamphetamine, 70% of patients stopped injecting and an additional 27% cut back; for methadone the corresponding figures were 67% and 21%. Both groups typically stayed in treatment for 10 months. Four psychotic episodes were associated with temporary return to, or increased use of, illicit amphetamine, and these remitted once illegal use had ceased. Two of these patients had been diagnosed psychiatrically ill prior to referral. No client who kept entirely to their prescribed medication experienced psychotic symptoms, even those suffering psychiatric illnesses.

8 A small study in Portsmouth of 26 amphetamine injectors prescribed 30mg daily amphetamine elixir on a supervised consumption basis also found no consequent serious mental health problems, though among a caseload selected to have none to begin with. According to clinical observations and the patients' responses to a confidential survey, there were considerable benefits, including stopping or reducing injecting and risky sharing of injecting equipment. Illegal amphetamine use and crime rarely stopped altogether but were substantially reduced over an average 15 months in treatment. Experience with these initial patients led the service to offer more intensive psychosocial support along with a higher-dose but time-limited (14 months) amphetamine prescribing programme, the intention apparently being to engineer more thorough lifestyle changes rather than to focus on retention in treatment.

9 Clinicians at a similar programme in Cardiff also found it necessary to increase the initial 30mg dose of dexamphetamine to a maximum of 60mg. Of 13 amphetamine injectors, two stopped attending early in treatment and one was withdrawn after becoming psychotic while still heavily using street amphetamine. For the remaining 10 participants, after 24 weeks the average number of injections had decreased from 38 a week to 1.3, and consumption of street amphetamine had decreased from 40gm a week to 1.6gm. Of the six former sex workers, five had stopped sex work completely and one only worked occasionally. All patients reported a reduction in criminal activity and an improvement in general health, an improvement also commented on by workers at the clinic not involved in the project. As in Portsmouth, mentally ill patients were not accepted for treatment.

10 As described in a review, a study in Manchester Merrill J., Tetlow V.A. "Prescribing for amphetamine users: evaluation of a dexamphetamine substitution programme." Paper presented at Beyond oral methadone: expanded pharmacotherapies for the treatment of drug dependence, New York, 1998. used urinalysis to confirm the accounts given by 50 amphetamine users prescribed dexamphetamine. After 12 months, four patients had achieved abstinence, 27 continued to be prescribed, and 19 had dropped out of treatment, 13 of whom had not used illicit amphetamine for over three months. Frequency of injecting and quantity of amphetamine used each week decreased over 90% during treatment. Response to treatment was rapid, with almost all of the improvement being evident within the first two or three weeks.

11 A study not limited to substitute treatment recruited 58 new amphetamine using clients identified by staff from 16 drug agencies in north-west England. All were daily or very frequent amphetamine users (on average using nearly six days a week) and 59% injected the drug. Average consumption was nearly 4gm a day. Researchers interviewed them within about a month of starting treatment and then roughly monthly for the next two months. At the last assessment 49 were left in the sample.

The impact of the treatment they received – for some prescribed dexamphetamine and others mainly counselling and advice – was assessed by comparing clients against individually matched controls as alike as possible, except that they had not entered treatment. Two months after treatment entry, 43% of clients no longer took illicit amphetamines, a rare development among controls. Clients had cut their daily consumption by about twice as much as their matched controls and used twice a week compared to 4–5 days a week. Among the injectors, over a third of the clients but only 4% of controls had stopped injecting. After treatment, 17% of clients reported committing non-drug crimes in the previous month compared to 67% in the three months before treatment. Among controls the corresponding figures were 93% and 50%. Fewer prescribed than non-prescribed clients continued to commit crimes and prescribing was associated with greater reductions in average daily amphetamine consumption and in injecting. However, prescribed clients were no more likely to become abstinent or to use amphetamines less frequently. Sympathetic counselling emerged as at least as important as prescribing.

Because the controls were not (as far as was known) trying to stop using amphetamines, the study could not assess what proportion of treatment clients might have succeeded in overcoming their drug problems, even without professional help. In one UK study from around the same time, nearly a third of attempts to self-detoxify from amphetamine were successful for at least three months, compared to 37–45% in the four to six months after treatment started in the current study.

International experience

One US reviewer has noted that "reticence" about developing substitute prescribing regimens for stimulants has held back such treatments despite their promise, while the search for agents to block stimulant use has not resulted in an effective approved medication.

This reviewer usefully summarised the concerns about prescribing amphetamine to amphetamine-dependent patients, and judged them "likely manageable" and less than the risks of untreated stimulant dependence, as long as good psychosocial therapy was available and there was appropriate monitoring. Among the concerns were increased blood pressure and heart rate. In practice these have not proved a major problem and can easily be monitored. Nevertheless, patients with a history of cardiovascular disease are best carefully screened and a risk-benefit determination made with respect to likely compliance and stable dosing v. patterns of abuse. Lesser, but still real concerns exist for thermoregulation with potent stimulant administration. This problem is nominal, infrequent, and can be readily addressed by clinical management. Potent stimulants, including amphetamines, were long used for weight control, but this effect rapidly wears off with repeated use, and patients can be expected to be healthier and better fed while stabilised in treatment than while using amphetamines out of treatment, an expectation confirmed in studies which have not found weight loss a problem. Psychiatric consequences of amphetamine at high doses are well known, including paranoia and hallucinations, but there is no indication of pervasive problems in any of the Australian, US, or UK reports. In prescribing trials it seems these events are uncommon, and no psychotic episodes have been reported among treatment-compliant patients who had not already experienced them. Rather, this has been restricted to participants who continued use of high-dose illicit stimulants, some of whom also had pre-existing psychotic episodes. Preventing diversion of prescribed supplies on to the illicit market is another concern dealt with in various ways including supervised consumption. When take-home doses have been allowed, urine testing and the supervision provided by family or others help ensure compliance. However, some degree of diversion may have to be accepted as the price of the benefits of prescribing.

Sydney in Australia was the location of the first trial to rigorously test dexamphetamine prescribing for amphetamine dependence by randomly allocating patients to this treatment plus counselling, or to counselling alone. Oral dexamphetamine was prescribed in doses up to 60mg daily and consumed under supervision for 12 weeks, including a final two weeks' withdrawal. The 41 patients were selected to be free of major physical or psychiatric problems, had used amphetamine on average for 10 years, and at treatment entry were using twice a day, nearly always by injection.

Small samples and modest effects meant there were no statistically significant differences in outcomes. The differences that there were favoured the prescribed group, in terms of use of illegal amphetamine, spending on these drugs, injecting, and infection risk behaviour. No psychotic symptoms were reported for any subjects. Prescribed patients on average attended three quarters of the time for their doses. Counselling was poorly attended, but significantly less poorly by the prescribed patients; 43% either declined counselling or failed to attend appointments compared to 75% of non-prescribed patients. This means that the trial was largely a comparison between prescribed amphetamine plus some counselling versus no treatment at all. The results seem indicative (as other studies have found) of the poor pulling power of counselling alone, and the ability of substitute prescribing to work synergistically with counselling by improving attendance. However, at 10 weeks, this trial cannot be considered a test of what most people would consider a maintenance programme as opposed to stabilisation prior to withdrawal.

The researcher who led the trial described above has summarised the evidence to 2002 and provided a useful account of the history of the treatment and its potential benefits and drawbacks. His advice is that substitution therapy may only be appropriate for severely dependent amphetamine users, most probably daily, injecting users, although this should not preclude heavily using non-injectors. The most enduring treatment benefits are likely to be achieved by a combination of pharmacotherapy with psychosocial intervention; prescribing attracts, stabilises and retains, while the therapy helps extend and embed lifestyle change preparatory to discharge, which seems to occur more often and earlier than in opiate substitute prescribing programmes.

Given the risk of precipitating a psychotic episode, the author saw screening and monitoring for psychotic symptoms as essential, and noted that some clinicians do not consider the treatment suitable for patients with a history of schizophrenia or bipolar affective disorder. Partly to avoid diversion on to the illicit market, supervised consumption, urinalysis and regular medical monitoring for side effects at least during the first three-month stabilisation period provides a fully informed basis for assessing the individual effectiveness of substitution and the appropriateness of continued maintenance. Unsupervised doses may be appropriate for stabilised patients where medication compliance indicates that diversion is unlikely. Patient commitment to supervised treatment protocols may also protect programmes from unsuitable or poorly motivated patients who may merely wish to obtain additional sources of drugs.

Other maintenance medications

Modafinil has been trialled as an alternative to amphetamine maintenance. The drug is a wakefulness-promoting agent prescribed to treat narcolepsy and other sleep problems. The drug seems promising as an amphetamine substitute because it counteracts fatigue and concentration difficulties common after amphetamine use stops and may help correct sleep patterns. It does not cause euphoria or tolerance so there is relatively little risk of patients becoming severely dependent on it (making it suitable for time-limited treatments) or of its diversion to the illicit market.

The first randomised trial with methamphetamine-dependent patients took place in Australia and compared 10 weeks of modafinil against placebo. The modafinil patients tended on some measures to make slightly greater reductions in their illegal stimulant use, but not to a statistically significant degree, and there was no extra impact on craving for methamphetamine or severity of dependence. Neither did the active medication promote engagement or retention in treatment. Modafinil appeared safe and well tolerated; there were no medication-related serious adverse events. The modafinil group did report more transient mild/moderate headaches and also tended to report more nausea/gastric discomfort, all of which resolved after the first week of treatment. Reports of thought disorder and anxiety were unique to the modafinil group and indicate caution in patients with pre-existing psychotic symptoms or anxiety.

The medication has also been tried with a small group of HIV positive gay men in New York, about half of whom substantially reduced their methamphetamine use, but there was no comparison group of patients against which to test whether the medication was responsible for the improvements.

Another US trial which recruited amphetamine-dependent patients and allocated them randomly to three treatments, found that compared with placebo, slow-release methylphenidate was associated with a statistically significant reduction in amphetamine injecting, but aripiprazole with the reverse. The authors concluded that effective pharmacological maintenance treatments for drug injectors must induce at least some euphoria, such as methadone and buprenorphine in opioid dependence, or methylphenidate in amphetamine dependence.

For an account of other non-maintenance medications trialled to reduce or block methamphetamine use see this report.

Comment on these background notes Return to main entry
L10 Web Stats Reporter 3.15 LevelTen Hit Counter - Free PHP Web Analytics Script
LevelTen dallas web development firm - website design, flash, graphics & marketing