|

 

Cns drugs 14: 329-341, nov 2000

CNS Drugs 2000 Nov; 14 (5): 329-341 Adis International Limited. All rights reserved.
Does Disulfiram Help to Prevent
Relapse in Alcohol Abuse?
Colin Brewer,1 Robert J. Meyers2 and Jon Johnsen3 1 The Stapleford Centre, London, England2 University of New Mexico, Albuquerque, New Mexico, USA3 Psykiatrisk Ungdomsteam, Asker og Bærum, Sandvika, Norway When taken in an adequate dose, disulfiram usually deters the drinking of alcohol by the threat or experience of an unpleasant reaction. However, unlessits consumption is carefully supervised by a third party as part of the formal orim-plied therapeutic contract, it is usually discontinued and the deterrent effectis therefore lost. In most studies, disulfiram administration has not been super-vised and most reviews fail to stress the crucial importance of supervision. Un-supervised disulfiram has little or no specific effect. We have therefore reviewedall published clinical studies in which there was evidence that attempts had beenmade to ensure that disulfiram administration was directly supervised at leastonce a week. We found 13 controlled and 5 uncontrolled studies. All but onestudy reported positive findings, which were usually both statistically and clini-cally significant in controlled evaluations. In the sole exception, involving ‘skid-row alcoholics', it seems that adequate supervision was not achieved. In general,the better the supervision, the better the outcome.
Provided that attention is paid to the details of supervision and that supervisors are given appropriate training, supervised disulfiram is a simple and effectiveaddition to psychosocial treatment programmes. Compared with unsuperviseddisulfiram or no disulfiram control groups, it reduces drinking, prolongs remis-sions, improves treatment retention and facilitates compliance with psychosocialinterventions such as community reinforcement, marital and network therapies.
The supervisor may be a health professional, workmate, probation officer orhostel worker but is usually a family member. Treatment should probably con-tinue for a minimum of 12 months. Supervised disulfiram appears to be moreeffective than supervised naltrexone and may be more effective than unsupervisedacamprosate. The crucial importance of supervising the consumption of disul-firam has been overlooked or minimised by many reviewers.
Brewer et al. 1. Disulfiram in Alcohol Abuse
intake. The importance of supervision was recog-nised by a few authors even in the early days of It has been well known for over 50 years that disulfiram treatment.[3,4] Numerous studies (discus- most patients who take disulfiram at an adequate sed in section 3) show that such third party involve- dosage will experience an unpleasant and occa- ment greatly improves compliance and therefore sionally dangerous reaction within a few minutes greatly improves the effectiveness of disulfiram. In of drinking alcohol. As little as half a unit of alco- 1986, an influential publication by the Royal Col- hol (approximately 5g) may be sufficient to cause lege of Psychiatrists[5] noted that ‘it is becoming the disulfiram-alcohol reaction (DAR).[1] Thus, di- more frequent for the doctor to suggest that a third sulfiram deters such patients from drinking alco- person supervises the [disulfiram] . . a relative or hol, or from repeating the experience if they have someone at work'.
already had an unpleasant DAR. Obviously, disul-firam has no effect on drinking behaviour if pa- 2. Early Reviews of Efficacy
tients for whom it is prescribed either discontinueit or do not take it in the first place.
Opinions have been divided over the effective- Some patients need to take higher dosages of di- ness of disulfiram in treating alcohol abuse. Fol- sulfiram than the usual range of 200 to 500 mg/day lowing a review of the literature, Soyka[6] conclu- in order to obtain plasma concentrations of the ac- ded that ‘Disulfiram . . lithium and various other tive metabolite of disulfiram sufficient to inacti- substances have been tested in an attempt to in- vate liver aldehyde dehydrogenase (ALDH). Other crease abstinence rates in alcoholic patients, all with patients are incapable of producing the active me- little or no success'. The reference he gives for this tabolite at concentrations high enough to inactivate statement, Fuller et al.,[7] is to a fairly classically the ALDH isoenzymes adequately. Thus, in some designed randomised controlled trial. Disulfiram patients disulfiram treatment will not cause a reac- presents certain difficulties in experimental design tion if alcohol is consumed.[2] which do not apply to most other drugs. Because Since many individuals with alcoholism are very the DAR can be dangerous, it would obviously be ambivalent about altering their drinking habits, it hazardous and unethical to inform the patients that is understandable that many of them will also be am- half of them would be taking a placebo, since they bivalent about taking disulfiram. This ambivalence might then be tempted to risk drinking with serious is not, of course, unique to disulfiram. Compliance consequences. Fuller et al.[7] got round this prob- rates with many treatments are surprisingly low, lem ingeniously. A third of the patients were pre- but good compliance is obviously of particular im- scribed oral disulfiram 250 mg/day, even though portance in individuals with alcoholism. For them, this dosage would not have been enough to produce as for patients with chronic schizophrenia, poor a sufficiently deterrent DAR in many patients.[1] A compliance with the medication which is pre- second group were told that they were receiving scribed specifically to treat their condition is not disulfiram but were only given 1 mg/day – a dosage just a common problem but can actually be an in- certainly insufficient to produce a DAR. The re- herent characteristic of the condition.
maining third received only riboflavine.
In the case of patients with chronic schizophre- At the time of the study by Fuller et al.,[7] unsu- nia, compliance can be improved by using depot pervised disulfiram treatment was standard prac- injections. Because there is no readily available tice for treatment of alcohol abuse in the US.[8] The pharmacologically effective depot preparation of study by Fuller et al.[7] was designed to rigorously disulfiram, other methods of improving compli- test the effectiveness of unsupervised disulfiram.
ance have to be used. These usually involve recruit- Therefore, as was subsequently pointed out,[9] al- ing a third party, such as a family member or a pro- though all patients were offered (and many re- bation officer, to supervise or monitor disulfiram ceived) follow-up counselling at weekly intervals  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Disulfiram in Alcohol Abuse Relapse Prevention for several months, provision was not made to en- their review, disulfiram was placed firmly in the sure that patients took at least 1 weekly dose of disul- latter category.[14] However, in more recent reviews, firam under supervision. However, the number of Miller[15,16] has revised this view and now regards abstinent days did increase slightly at 12 months in supervised (but not unsupervised) disulfiram as a the group receiving disulfiram 250 mg/day. The work treatment of proven effectiveness. In recent meta- of Azrin et al.[10] showed that nearly all patients analyses of the literature,[17-19] 2 of the treatment prescribed disulfiram without third party supervi- approaches for alcohol abuse which consistently sion had discontinued it within 3 months. The small perform well are Community Reinforcement Ther- proportion of patients who regularly take disul- apy (CRT) [see section 3] and Behavioural Marital firam even without supervision – about 20% in thestudy by Fuller et al.[7] – thus appear to be a very Therapy, both of which lend themselves to (and atypical and unusually compliant group of patients.
often incorporate) supervised disulfiram therapy.[20] They may be similar to the patients in a study by It is interesting that in his review[16] of studies Edwards et al.[11] nearly 50% of whom had good of treatment programmes including disulfiram, outcomes at 12 months despite having no active Miller includes studies of disulfiram implants.
treatment of any kind following an elaborate initial Studies involving disulfiram implants have gener- research-oriented assessment. In a letter published ally produced better results than trials involving after the publication of the trial, Fuller[8] noted the patients treated with unsupervised oral disulfiram.[21] importance of supervision and reported that he had However, any effectiveness cannot be attributed to attempted, unsuccessfully, to get funding for a fur- a pharmacological process since Johnsen and Mor- ther study in which the contribution of adequate di- land[21] have shown conclusively that for commer- sulfiram supervision would be separately assessed.
cially available disulfiram implants, it is impossi- Unfortunately, as so often happens after the pub- ble to detect a blood concentration of disulfiram lication of influential papers, subsequent criticism, (or its presumed active metabolites) and that alco- published as letters, is often ignored by later review-ers, even if it is accurate.[12] Although the study by hol administered intravenously under blinded con- Fuller et al.[7] is often discussed in reviews because ditions, does not provoke a DAR. All pharmacolog- of its thorough design, the limitations of a study of ical treatments have nonspecific or placebo effects unsupervised disulfiram should be considered by as well as pharmacological effects. Disulfiram is no reviewers. Reviewers need to distinguish between the mainly negative results of a larger number of It may be noted at this point that disulfiram was trials of unsupervised disulfiram, which, in our view, generally preferred over the only alternative alcohol- often have poor designs, and the much smaller sensitising drug, calcium cyanamide (now no longer number of trials in which the consumption of di- available), because of the relatively short half-life sulfiram is more or less diligently supervised. Of of the latter – about 12 hours. This made the task of these latter studies, the study by Azrin et al.,[10] which supervising oral medication much more demand- produced the most positive results of all such trials, ing than with disulfiram, whose alcohol-sensitising was noted in a review by Saunders[13] as having a effects will usually persist for at least 2 or 3 days particularly convincing experimental design.
after the last dose and may sometimes last for up to In a review by Miller and Hester[14] in 1986, treatments for alcoholism were classified into those a week or even more. Liskow et al.[22] found that for which sound evidence for effectiveness exists; when patients drank after discontinuing disulfiram, those whose effectiveness or specific effectiveness the time between the last dose of disulfiram and has yet to be demonstrated; and those which are drinking averaged 51 ± 50h on the first occasion, demonstrably lacking in any specific effect. In 60 ± 58h on the second, and 52 ± 52 h on the third.
 Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Brewer et al. 3. A Review of Studies of
poor prognosis ‘skid-row alcoholics'. In this study, disulfiram was combined in a capsule with chlordi-azepoxide at a dosage of 75 to 125 mg/day. The idea To assess whether disulfiram successfully pre- was that chlordiazepoxide would act as a positive vents relapse in alcohol abuse, we reviewed the reinforcer which would be an incentive for patients literature on disulfiram in alcoholism treatment.
to continue taking disulfiram, rather as in a later Because of evidence that unsupervised disulfiram study (Liebson et al.[29]) combining disulfiram with is of little value, clinical studies were included only methadone maintenance therapy. Depending on if there was evidence that attempts had been made the time of administration, the chlordiazepoxide to ensure that disulfiram administration was direct- would act as either a tranquilliser or a hypnotic – ly supervised at least once a week. A search of MED- effects which would be attractive to many patients LINE up to January 2000, previous reviews and with alcoholism and much less damaging than us- any other papers known to the authors revealed 13 ing alcohol for these purposes. Weekly supervision controlled and 5 uncontrolled studies. These are re- is implied, though not entirely clear, but improving viewed chronologically and are summarised in ta- compliance was clearly central to the study. This was a pilot study but 6 patients had remained in Bourne et al.[23] published the first study in which treatment for a mean of 6 months at the time of the disulfiram was routinely supervised, generally as one component of a probation order. Although the In a retrospective study conducted in Colorado study was uncontrolled, the results were very en- Springs, US, Haynes[26] investigated the effective- couraging given that virtually all these patients were ness of supervised disulfiram for 12 months as one offenders with recurrent alcoholism with long his- condition of a probation order in 138 offenders with tories of severe alcohol abuse resistant to other recurrent alcoholism. Some patients left town, of- treatment methods. About 60% of the 196 study par-ticipants were compliant with disulfiram treatment ten for legitimate reasons, and 12% were jailed for and were supervised by probation officers, during noncompliance. In the remainder, acting as their the 30 to 60 days of a suspended prison sentence.
own controls, after 12 months there was an almost Most took it for longer than was legally required.
13-fold reduction in alcohol-related offenses com- The authors came to the conclusion that probation- pared with the participants' previous record.
linked supervised disulfiram seemed to be a useful The first published study which investigated ob- idea and worth developing.
jectively the relationship between supervision and Gallant et al.[24] undertook a randomised, con- outcome was done by Gerrein et al.[27] in 49 patients.
trolled study of compulsory versus voluntary treat- There was a significantly better outcome when ment of 84 offenders with chronic alcoholism which disulfiram treatment was supervised for 8 weeks theoretically included disulfiram administered un- during daily outpatient attendance compared with der supervision. Unfortunately, it seems that pa- unsupervised disulfiram treatment.
tients very rarely turned up for treatment and there Azrin[28] published the first of 2 studies in which were no immediate sanctions for noncompliance.
he investigated the effects of both supervised and Accordingly, very few patients seem to have actu- unsupervised disulfiram combined with CRT, a ally received disulfiram. We question if this study package of essentially behavioural (as opposed to should be regarded as a valid assessment of super- psychodynamic) outpatient interventions. CRT vised disulfiram treatment, but we have included it had already been shown in a study by Hunt and for the sake of completeness. It is the only study to Azrin[40] to be significantly more effective than have found no benefit.
conventional outpatient treatment, and both CRT Liebson and Faillace[25] described an ingenious and the methodology employed by Azrin[28] have of- method of improving compliance in a group of 10 ten been mentioned as examples of good practice in  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Table I. Clinical trials of supervised disulfiram
patients disulfiram supervision groupa Bourne et al.[23] Skid-row alcoholics Compliance approximately 60% Gallant et al.[24] Recurrent skid-row alcoholic Few patients in any group attended Skid-row alcoholics Disulfiram combined in capsule with Not stated Recurrent alcoholic offenders 13-fold reduction in arrests Gerrein et al.[27] Improved longer term retention Unsupervised ?250 Better results with less counselling. 2 year Liebson et al.[29] Unsupervised 250-500 Patients with alcoholism 1 vs 17% drinking days receiving methadone maintenance therapy Robichaud et al.[30] RCT Daily/alternate days Employees with alcoholism 5-fold reduction in absenteeism Daily supervisor with Rural outpatients Nearly 100% abstinence at 6 months in Offenders with recurrent Average abstinence 30 weeks, vs 6 weeks in previous 2 years. 9/16 totally successful Not stated Outpatients Contracting improved compliance and Sereny et al.[33] Outpatients where treatment 40% total success; 18% partial success had failed on 3 previous p ≤0.01 on some measures ascorbic acid(vitamin C) Gerber et al.[35] Not stated Outpatients with liver disease Quality of life and liver function normalised Besson et al.[36] Not stated Outpatients Disulfiram improved acamprosate effects Carroll et al.[37] Patients who abused alcohol Reduced drinking associated with reduced Carroll et al.[38] Twice weekly/weekly Patients who abused alcohol Longer retention in treatment (p < 0.05) Tønnesen et al.[39] RCT Outpatients with alcoholism Complete preoperative abstinence in In most cases, the control group received at least standard levels of psychosocial treatment.
Main outcome measures.
Disulfiram patients unrandomised in a RCT of acamprosate.
NA = not applicable; NS = not significant; RCT = randomised controlled trial; U = uncontrolled.
Brewer et al. treatment and research (e.g. Saunders[13]). Disul- supervision involved both associates and counsel- firam treatment was supervised in the disulfiram lors. One of us (also one of Azrin's co-authors) has group (i.e. 50% of patients) both by family mem- recently tried to replicate this study with William bers and by counsellors at each counselling ses- Miller. The work is not yet complete but it proved sion. Apart from the highly significant differences almost impossible to randomise patients to the di- in favour of supervised disulfiram, this study is sulfiram group because of the inability of the proj- noteworthy for the unusually long follow-up pe- ect to obtain immediate evaluation for disulfiram pre- riod (2 years) during which the improvements were scribing. Many patients had to wait up to 4 weeks to actually receive disulfiram. Consequently, the Az- Liebson et al.[29] studied supervised disulfiram in rin protocol could not be truly replicated.
6 patients receiving methadone maintenance ther- Brewer and Smith[31] published a pilot study of apy who also abused alcohol. In the experimental 16 offenders with chronic alcoholism who were at- group, dispensing of the daily methadone dose was tending London courts with an average of 6.3 alco- made contingent on taking disulfiram under profes- hol-related convictions and an average maximum sional supervision, and alcohol consumption was period of abstinence outside prison of only 6 weeks.
considerably and significantly lower than in the un- Patients were offered regular counselling and su- supervised control group.
pervised disulfiram treatment at the probation of- Robichaud et al.[30] also found a significant im- fice as conditions of probation. At the end of the provement when supervised disulfiram was used as study, the average maximum period of abstinence virtually the sole treatment in an employee alcohol- for the whole group was 30 weeks and all but one ism programme. The 21 patients in the study were re- participant had exceeded their longest abstinence quired to take disulfiram at work under nursing super- in the previous 2 years.
vision for an average of 10 months as a condition Keane et al.[32] used ‘spouse contracting' in a of remaining in employment. The absenteeism rate study of 25 patients to try to improve disulfiram before treatment was 9.8%. During disulfiram treat- compliance. The control group were simply encour- ment, it fell to 1.78% and rose again to 6.7% when aged to use disulfiram. The contract group did the disulfiram was discontinued. Counselling was somewhat (but not significantly) better during the 3 also offered to these patients, most of whom had pre- month study but unlike the studies by Azrin[28] and viously had alcoholism treatment, but few of them Azrin et al.[10] there was no professional supervi- took up the offer.
sion of disulfiram consumption during counselling To follow up his earlier study, Azrin and his co-workers conducted another study in 43 patients A prospective study by Sereny et al.,[33] though which confirmed the effectiveness of properly su- not controlled in the classic fashion, gave rather pervised disulfiram.[10] The study also made the impressive results (see below). Noting that a sig- very important (and unexpected) discovery that for nificant number of patients relapsed repeatedly de- patients with reasonably intact relationships, who spite compliance with a conventional treatment pro- constitute, in many studies, a majority or at least a gramme, they devised a radical but constructive large minority of participants, involving the part- response to patients who had relapsed at least 3 times.
ners who do not have alcoholism and giving them Instead of declining to offer further treatment, they simple training to improve the quality of supervi- told them that they would be accepted for further sion was the most important component of treat- treatment but only if they agreed to take disulfiram ment. In such cases, adding more intensive coun- under professional supervision during their outpa- selling conferred no additional benefit. Disulfiram tient attendance. 68 of 73 patients agreed to this effects were maintained throughout the 6 months arrangement. In this study ‘total success' was defin- study period. As in the previous study by Azrin[28] ed as being sober for at least 6 months and remain-  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Disulfiram in Alcohol Abuse Relapse Prevention ing in the mandatory disulfiram programme at the amyl transpeptidase levels, and mean corpuscular time of assessment, or having been discharged from volume all returned to normal in this period.
mandatory disulfiram after 12 months of sobriety.
Some studies have included new pharmacother- By these criteria, 27 patients (40%) were totally suc- apies such as acamprosate and naltrexone. An ex- cessful. 12 patients (18%) were partially success- ample is the controlled study involving 46 patients ful, treatment failed in 20 (29%) and treatment out- of disulfiram and acamprosate by Besson et al.[36] come was undetermined in 9 (13%). These are good which produced interesting results. An important results in a group of patients who by definition would conclusion of the study is that the effectiveness of normally be regarded as having a poor prognosis.
acamprosate is increased by combining it with di- In other respects, the management of these patients sulfiram given under professional supervision. Pa- appears to have been similar to that provided on pre- tients were randomised to acamprosate but not to vious occasions. The good outcome, compared with disulfiram. However, a breakdown of the results shows both the previous outcomes in these patients and that the patients who took disulfiram alone had a that of other patients typically treated in the same better outcome and higher retention than those who centre without supervised disulfiram, thus seems took acamprosate alone.
likely to have been attributable to the addition of The effectiveness of disulfiram can also be seen supervised disulfiram to the programme.
in a comparison of disulfiram with naltrexone in in- A multicentre study of 6 months' duration con- dividuals who abused both alcohol and cocaine.[37] ducted in the UK by Chick et al.[34] confirmed the In a 12 week study, 18 patients were randomised effectiveness of supervised disulfiram. This study to disulfiram or naltrexone, supervised weekly by was also designed to discover whether the effect- a nurse. Attrition was high in both groups but lower iveness was attributable to the psychological and with disulfiram. On all measures of both cocaine and symbolic impact of supervision or to the deterrent alcohol use, the disulfiram group did significantly and pharmacological effects of disulfiram. The 126 better than the naltrexone group.
participants were patients receiving standard out- A recent randomised controlled trial conducted patient treatment for alcoholism who were ran- by Carroll[38] of 122 patients who abused both al- domly assigned to supervised disulfiram or super- cohol and cocaine compared various types of psy- vised treatment with ascorbic acid (vitamin C).
chotherapy – cognitive-behavioural therapy (CBT) Where possible, supervision was delegated to fam- and twelve-step facilitation (TSF) – with ordinary ily members who were given appropriate instruc- clinical management (CM) with and without disul- tion but in other cases, medication was supervised firam. Disulfiram ingestion was monitored by a by clinic staff or community nurses. The results, nurse twice weekly for the first month of treatment which included a significant reduction in γ-glut- and weekly thereafter. According to the study au- amyl transpeptidase levels, very clearly favoured thors,[38] ‘The CBT/disulfiram group had the high- the disulfiram group.
est rate of retention (mean 8.8 weeks), followed by Gerber et al.[35] studied quality of life (QoL) and CM/disulfiram (8.4 weeks), TSF/disulfiram (8.0 liver function in a group of 20 patients with alcoh- weeks). Subjects assigned to disulfiram treatment olism receiving supervised disulfiram. QoL and were retained significantly longer than those assign- liver function were assessed at baseline and after 6 ed to no medication [8.4 versus 5.8 weeks (p < 0.05)].
months. QoL was also assessed in 20 volunteers No significant differences in retention by psycho- matched for gender, age, education and social sta- therapy were found'. It has recently been suggested tus. At baseline, but not after 6 months, patient that disulfiram has significant effects in reducing QoL was significantly lower than that of volun- cocaine use even in patients who do not have co- teers (ANOVA p < 0.01). Bilirubin levels, γ-glut- morbid alcohol abuse.[41]  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Brewer et al. Carroll et al.[38] also found significant effects of are a unique feature of treatment with disulfiram or disulfiram on consecutive weeks of cocaine absti- other alcohol-sensitising drugs.
nence, alcohol abstinence and abstinence from both First, there is the real and unpleasant DAR. Pa- cocaine and alcohol: ‘Effect sizes (d) for disulfiram tients taking disulfiram may know in various ways compared with no medication on duration of absti- about this reaction. All of them will have been warned nence from cocaine, alcohol and both were, respec- about it and many will believe what they have been tively, 0.42, 0.68 and 0.46'. In contrast, the specific told without the need to directly experience it them- effect of the psychotherapies was rather modest: selves. Some know of it vicariously from observing ‘Effect sizes for the active psychotherapies com- or hearing about the reaction in other individuals. A pared with CM on duration of abstinence were 0.16 variable proportion of patients make the discovery for for cocaine, 0.11 for alcohol and 0.18 for both co- themselves by actual experiment.[1] According to caine and alcohol'.
Liskow et al.[22] over 75% of those who drank while This is an important study, not only because it receiving disulfiram reported experiencing a DAR, is further evidence for the effectiveness of super- with more than one-third of those experiencing the vised disulfiram, but also because retention in ther- DAR reporting it to be severe (85% of patients were apy is desirable in many cases, though probably not taking 250 mg/day, the rest 500 mg/day). It is only in all. Clinicians working in relapse prevention ob- because disulfiram has, and is known to have, this viously need to spend enough therapeutic time with very real potential for an aversive effect that it deters patients to help them to make positive cognitive many patients from drinking by the mere fact of tak- and behavioural changes. In our view, supervised ing it. Further evidence for a specific deterrent effect disulfiram is clearly one of the most effective tech- of the DAR comes from Japanese studies showing that niques for maximising treatment retention but the per capita alcohol consumption is reduced in those sobriety that it usually imposes also gives patients parts of Japan where there is a high incidence of inac- a better chance of dealing with ambivalence and tive forms of ALDH. Other studies show that Asians denial and of learning and consolidating new cop- in whom 1 parent (one allele) has the inactive ALDH ing skills and strategies.
are protected from 75 to 90% against alcoholism and Finally, Tønnesen et al.[39] used twice-weekly sup- in these individuals, heavy drinking is reduced by ervised disulfiram as the sole treatment in a control- 66% even if they were born and raised in the US or led study of intervention versus no intervention in Canada.[42,43] Furthermore, individuals who inherit 42 patients who were drinking more than 60g of an inactive ALDH from both parents are complete alcohol daily to measure the benefits of abstinence for a month before major surgery. All patients re- Secondly, taking disulfiram regularly (or having ceiving disulfiram apparently abstained com- an implant inserted with a supposed active life of 3 pletely and had significantly fewer postoperative or 6 months) surely has certain symbolic connota- complications (31 vs 74%, p < 0.02) than the non- tions. It indicates that here is a patient who is will- intervention group.
ing, however uncertainly or ambivalently, to surren-der some control over his or her freedom or urge to 4. Components of Effectiveness in
drink. Such patients announce both to themselves and to the wider world that they are not merelytalking about changing their drinking habits, or The powerful deterrent effect that has been seen making often unconvincing promises to do so, but in some studies of disulfiram implants (e.g. Johnsen are actually doing something about it. These pa- and Morland[21]) which actually have no measurable tients are at the ‘action' stage in the well known pharmacological activity underlines the importance Prochaska and DiClemente model of changing ad- of 3 separate but mutually reinforcing factors which dictive behaviour. Furthermore, the patient is in-  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Disulfiram in Alcohol Abuse Relapse Prevention volving some third party (the family member or external control and monitoring is crucial to max- probation officer in the case of oral medication; the imising the benefits of supervised disulfiram. There surgeon in the case of the implant) for the specific is evidence from studies of anxiety disorders that purpose of strengthening a resolve which he or she psychological treatments may improve medication knows is often tenuous, varies from one day to an- compliance.[44] Support, encouragement and ex- other, or not infrequently fails altogether.
planations can imbue confidence in patients con- Finally, the involvement of a third party in super- cerning the efficacy of the prescribed medication.[44] vising oral disulfiram provides additional oppor-tunities for involving family members in the broader 4.1 The Technique of Supervision therapeutic and monitoring enterprise. Any failureof compliance is thus more likely to be detected Having presented the case for the importance of and reported promptly enough for professionals to supervising the consumption of disulfiram rather intervene, either before drinking resumes or before than simply leaving it up to the patient, let us now a mere lapse turns into a full-blown relapse. A large examine the process of supervision. It sounds, and proportion of disulfiram users (33% on the first is, a simple enough concept but as with many sim- occasion, 43% on the second, and 48% on the third ple procedures, such as giving an intramuscular in- occasion of disulfiram use) gave ‘desire to drink' jection, measuring the blood pressure, or taking the as their reason for stopping disulfiram.[22] Accord- temperature with an oral thermometer, there are ing to Liskow et al.,[22]‘This suggests that when pat- right and wrong ways of doing it and therefore at- ients and treatment personnel discuss whether di- tention to detail is important. Even in those studies sulfiram should be discontinued, this reason should where the importance of supervision is recognised, be explored vigorously'. An awareness of this po- few spell out the process in detail. Azrin[45] and tent combination of pharmacology, symbolism and Chick et al.[34] are exceptions. Table II summarises Table II. Recommended approach to the supervision of disulfiram treatment (reproduced from Azrin,[45] with permission)
Identify a disulfiram monitor who would be substantially and negatively affected by resumption of drinking, e.g. spouse, family member,employer, partner, landlord The monitor should normally have regular, ideally daily, contact with the patient Specify precisely the time and place where the disulfiram could be taken conveniently, with both persons present Have disulfiram taken at a time when other forms of medication are normally taken, i.e. the ‘response-chaining' principle Grind up the disulfiram tablet and dissolve it in a drink (coffee, tea, juice) to avoid any suspicion of later expulsion If the monitor is not present when the patient has taken the disulfiram, the patient should take another tablet the same day, when themonitor is present, to provide absolute assurance to the monitor The patient should thank the monitor for taking the time to observe The monitor should comment on some positive attribute of the patient, that is associated with sobriety, i.e. job status, love by children,doing jobs around the house, financial security At each therapeutic session, the monitor attends with the patient, if possible, so that the therapist can instruct, supervise, and providefeedback to both At each therapeutic session, the disulfiram is taken in the presence of the therapist The monitor is to telephone the therapist if the patient omits taking disulfiram for 3 days; the therapist then telephones the patient toarrange a session When the usual 30-day supply of tablets is nearly depleted, the monitor prompts and assists the patient to renew the prescription; failureto do so has been one of the most apparent major causes of discontinuing disulfiram The therapist asks the patient and monitor to rehearse probable situations which cause the reluctance to take the disulfiram, and teachesthem how to overcome such interferences The patient is taught to view the use and ritual of taking disulfiram as a means of providing assurance to themselves and their loved onesthat they will not succumb to temptations that are otherwise beyond their control. It is emphasised that the central feature is the patient'sdesire, not coercion  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Brewer et al. detailed but fundamentally simple and sensible ad- cluded it from large trials such as Project Match.[49] vice based on the study by Azrin.[45] Galanter,[46] whose ‘network therapy' makes exten- The involvement of families and other individ- sive use of disulfiram, supervised by a network of uals or institutions in the treatment process is cen- family members, friends or colleagues, stated that tral to the concept of CRT and of the Network Ther- that 12 months was a reasonable minimum duration apy described by Galanter.[46] Both these models of treatment. According to Galanter,[46] abstinence recognise that many patients with alcoholism ap- is often ‘well established' by then but he also states pear to need and benefit from psychosocial inter- that some patients wish (or should be advised) to ventions but also that any benefit is likely to be take it for longer. There are several case reports of minimised or lost altogether if patients do not re- patients who evidently felt they needed the protec- main sober for most of the time. Patients who are tion of disulfiram for 10 or 15 years. At least 2 of intoxicated will not easily learn the new cognitive these involved supervised disulfiram.[50,51] Con- and behavioural skills that the psychosocial inter- versely, one certainly comes across patients who ventions are largely designed to teach them. Super- seem to remain abstinent or achieve controlled vised disulfiram therefore seems likely to facilitate drinking for many years after only a few months of psychosocial interventions because of increased supervised disulfiram. However, we feel that courses compliance with psychological therapy. It should of less than 6 months are likely to be too short for also reinforce their effect because if supervised di- most patients.
sulfiram deters a patient from having recourse toalcohol when he or she would normally (for what- ever reason) feel tempted to use it, then that patientis obliged to practice helpful alternatives to drink- Reviews even of such well established treatments ing, the inculcation of which is one of the main aims as antidepressant drugs sometimes require meta- of psychological components of treatment.[47,48] analysis to accommodate both the numerous neg- As the work of Azrin[28] demonstrated very clearly, ative reports and the positive majority. In our view, taking supervised disulfiram makes it much less no such statistical ingenuity is needed to reach a likely that treatment will be adversely affected be- conclusion about the effectiveness of supervised cause the patient has lost his or her job, has no where disulfiram. With only one exception, all the control- to live or has finally destroyed his or her marriage, led studies reviewed in this article demonstrate an because of yet another episode of intoxication.
improvement over a variety of differently treated Galanter[46] stresses that the role of the supervisor control groups which is not just statistically signif- is usually to encourage regular disulfiram con- icant but often large, obvious and clinically impor- sumption and to report noncompliance promptly, tant. In the one study of supervised disulfiram that rather than to be actively coercive. Fortunately, pa- did not show a positive effect,[24] the patients were tients will often do things for their therapist a group of skid-row alcoholics who, it seems, sim- (whether physician or psychologist) that they ply failed to turn up regularly for treatment.
would not do for themselves or for their partners.
However, in our view the importance of super- vision has not been given the recognition it de- 4.2 Duration of Treatment serves by some reviewers. On the subject of super-vised disulfiram use, Gatch and Lal[52] say that Long term outcome studies are notoriously dif- ‘more recent reviews have recommended that di- ficult and expensive to do and are often vitiated by sulfiram works best when used as part of a treatment high withdrawal rates. This is true of all treatment plan that includes careful monitoring, psychologi- modalities. Azrin[28] followed up a small cohort for cal therapy and social support . .'; this is the only 2 years but, in our view, the undeserved unpopu- mention in their review of the importance of super- larity of supervised disulfiram has generally ex- vision. In a recent review, Hughes and Cook[53]  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Disulfiram in Alcohol Abuse Relapse Prevention conclude that ‘supervised oral disulfiram in a com- complication of disulfiram treatment – fulminant prehensive treatment programme seems to have hepatitis – is probably related to nickel sensitivity, some efficacy in certain individuals'. We find it which may explain why, although very rare, the maj- disappointing that they did not incorporate this im- ority of cases have been reported in women.[60] Even portant conclusion into their abstract. Discussing severe alcoholic liver disease is not a contraindica- the highly positive results of the study by Liebson tion to disulfiram treatment[60] and major psychiatric et al.,[54] Hughes and Cook[53] say that the study illness is not an absolute contraindication.[61] In all demonstrated ‘the benefits of supervising disul- instances, the risks of disulfiram use must be weigh- firam treatment, rather than the efficacy of disul- ed against the high mortality and morbidity of un- firam per se'. However, as we have already noted, checked alcohol abuse.
if swallowing disulfiram did not make an unpleas- While we welcome a wider range of pharmaco- ant reaction with alcohol highly probable, then su- logical interventions for alcoholism, 1 of the 2 most pervising it would surely have little effect. Disap- recent additions – naltrexone – seems to be less ef- pointingly, in nearly half the controlled studies fective than supervised disulfiram.[37] The same may reviewed in their paper, disulfiram was either un- be true of acamprosate, though there has been no tru- supervised or rather badly supervised and the tech- ly comparative study. However, both may some- nique of supervision is not discussed. Like several times be usefully combined with disulfiram.
other reviews, there is quite a lot of discussion about A recent report[59] notes a 10-fold variation in the unsupervised study by Fuller et al.[7] in this re- disulfiram prescribing between the highest and view and there is no mention of the subsequent corr- lowest prescribing of 13 countries. Three Anglo-Saxon countries – the US, New Zealand and Brit- espondence. In the discussion of the positive results ain – have the lowest figures. Gunne[62] has noted of the study by Azrin et al.,[10] in which the study the prevalence of anti-medical (and often anti-sci- authors themselves stress the important role of su- ence) attitudes among some controllers and pro- pervised disulfiram, Hughes and Cook[53] put more viders of addiction treatment. If, as we believe, emphasis on the potential improvement in outcome these attitudes are particularly evident in the US, if unmarried individuals in the study had got mar- this may partly explain the neglect of an old but ried. In contrast, Litten et al.[55] stress the potential still useful treatment.
of disulfiram if attention is paid to compliance.
Although we do not suggest that supervised di- sulfiram is needed for all patients, there are severalsituations in which it seems particularly helpful.
We are very grateful for comments and advice from For example, patients with a history of repeated Professor Christopher Cook, Marc Galanter and Yedy Israel.
treatment failure (especially after nonpharmaco-logical treatments), patients who have many drink- ing triggers and those facing serious consequences 1. Brewer C. How effective is the standard dose of disulfiram? Br J Psychiatry 1984; 144: 200-2 if they relapse. As well as sobriety, the benefits com- 2. Johansson B. Pharmacological studies on disulfiram (antabuse) monly include reduction in family worry, increase [dissertation]. Lund University, Sweden, 1989 in family trust and involvement in treatment, a reduc- 3. Fox R. Antabuse as an adjunct to psychotherapy in alcoholism.
N Y State J Med 1958; 58: 1540-56 tion in demoralising ‘slips', improvements in self- 4. Billet SL. The use of antabuse: an approach that minimises fear.
confidence and self-image, and more opportunities Med Ann Dist Columbia 1964; 33: 612-4 to receive positive feedback from family members 5. Special Committee of the Royal College of Psychiatrists. Alco- hol: our favourite drug. London: Tavistock Publications, 1986: and friends.[56] These benefits far outweigh the rel- atively small risks of treatment.[57,58] In Britain, the 6. Soyka M. Relapse prevention in alcoholism, recent advances and future possibilities. CNS Drugs 1997; 4: 313-27 National Poisons Centre knows of hardly any deaths 7. Fuller R, Branchey L, Brightwell D, et al. Disulfiram treatment from the DAR.[59] The only other potentially lethal of alcoholism. JAMA 1986; 256: 1449-55  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Brewer et al. 8. Fuller R. Disulfiram treatment of alcoholism [letter]. JAMA 29. Liebson I, Bigelow G, Flamer R. Alcoholism among methadone patients: a specific treatment method. Am J Psychiatry 1973; 9. Brewer C. Disulfiram treatment for alcoholism [letter]. JAMA 30. Robichaud C, Strickler D, Bigelow G, et al. Disulfiram mainte- 10. Azrin NH, Sissons RW, Meyer RJ, et al. Alcoholism treatment nance employee alcoholism treatment: a three-phase evalua- by disulfiram and community reinforcement therapy. J Behav tion. Behav Res Ther 1979; 14: 618-21 Ther Exp Psychiat 1982; 13: 105-12 31. Brewer C, Smith J. Probation-linked supervised disulfiram in 11. Edwards G, Orford J, Egert S, et al. Alcoholism: a controlled the treatment of habitual drunken offenders: results of a pilot trial of ‘treatment' and ‘advice'. Q J Stud Alcohol 1977; 38: study. BMJ 1983; 287: 1282-3 32. Keane TM, Foy DW, Nunn B, et al. Spouse contracting to in- 12. Bhopal RS, Tonks A. The role of letters in reviewing research.
crease antabuse compliance in alcoholic veterans. J Clin Psy- BMJ 1994; 308: 158-223 chol 1984; 40: 340-4 13. Saunders B. Treatment does not work: some criteria of failure.
33. Sereny G, Sharma V, Holt J, et al. Mandatory supervised An- In: Heather N, Robertson I, Davies P, editors. The misuse of tabuse therapy in an out-patient alcoholism program: a study.
alcohol: crucial issues in dependence, treatment and preven- Alcohol Clin Exp Res 1986; 10: 290-2 tion. London: Croom Helm, 1985: 102-16 34. Chick J, Gough K, Falkowski W, et al. Disulfiram treatment of 14. Miller WR, Hester RK. The effectiveness of alcoholism treat- alcoholism. Br J Psychiatry 1992; 161: 84-9 ment methods: what research reveals. In: Miller WR, Heather 35. Gerber M, Joyce CRB, Christen A, et al. Disulfiram adminis- N, editors. Treating addictive behaviours: processes of change.
tered by a trustee improves quality of life (QoL) in alcoholic New York (NY): Plenum, 1986 outpatients [abstract]. Eur Psychiatry 1994; 9: 184 15. Miller WR. The effectiveness of alcoholism treatment modal- 36. Besson J, Aeby F, Kasas A, et al Combined efficacy of acam- ities: causes and consequences of alcohol abuse. Hearings prosate and disulfiram in the treatment of alcoholism: a con- before the committee on governmental affairs. Pt III: Wash- trolled study. Alcohol Clin Exp Res 1998; 22: 573-9 ington, DC: US Senate, 1989: 171-85 37. Carroll KM, Ziedonis D, O'Malley, et al. Pharmacologic inter- 16. Miller WR. The effectiveness of treatment for substance abuse: ventions for abusers of alcohol and cocaine: a pilot study of reasons for optimism. J Subst Abuse Treat 1992; 9: 93-102 disulfiram versus naltrexone. Am J Addict 1993; 2: 77-9 17. Holder H, Longabaugh R, Miller W, et al. The cost effectiveness 38. Carroll KM, Nich C, Ball SA, et al. Treatment of cocaine and of treatment for alcoholism: a first approximation. J Stud Al- alcohol dependence with psychotherapy and disulfiram. Ad- cohol 1991; 52: 517-40 diction 1998; 93 (5): 713-28 18. Finney JW, Monahan SC. The cost-effectiveness of treatment 39. Tønnesen H, Rosenberg J, Nielsen H, et al. Effect of preopera- for alcoholism: a second approximation. J Stud Alcohol 1996; tive abstinence on poor postoperative outcome in alcohol mis- users: randomised controlled trial. BMJ 1999; 318: 1311-6 19. O'Farrell TJ, Bayog RD. Antabuse contracts for married alco- 40. Hunt G, Azrin NH. A community reinforcement approach to holics and their spouses: a method to maintain Antabuse in- alcoholism. Behav Res Ther 1973; 11: 91-104 gestion and decrease conflict about drinking. J Subst Abuse 41. Petrakis IL, Carroll KM, Nich C, et al. Disulfiram treatment for Treat 1986; 3: 11-8 cocaine dependence in methadone-maintained opioid addicts.
20. O'Farrell TJ. Treating alcohol problems: marital and family Addiction 2000; 95: 219-28 interventions. New York (NY): Guilford Press, 1993: 170- 42. Harada S, Agarwal DP, Goedde HW, et al. Possible protective role against alcoholism for aldehyde dehydrogenase. Lancet 21. Johnsen J, Morland J. Depot preparations of disulfiram: exper- imental and clinical results. Acta Psychiatr Scand 1992; 86: 43. Tu G-C, Cao Q-N, Israel Y. Inhibition of gene expression by triple helix formation in hepatoma cells. J Biol Chem 1995; 22. Liskow B, Nickel E, Tunley N. Alcoholics' attitudes toward and experiences with disulfiram. Am J Drug Alchol Abuse 1990; 44. Lader MH, Bond AJ. Interaction of pharmacological and psy- chological treatment of anxiety. Br J Psychiatry 1998; 173: 23. Bourne PG, Alford JA, Bowcock JZ. Treatment of skid-row alcoholics with Disulfiram. Q J Stud Alcohol 1966; 27: 42-8 45. Azrin NH. Disulfiram and behaviour therapy: a social-biochem- 24. Gallant DM, Bishop MP Faulkner MA, et al. A comparative ical model of alcohol abuse and treatment. In: Brewer C, ed- evaluation of compulsory (group therapy and/or antabuse) and itor. Treatment options in addiction: medical management of voluntary treatment of the chronic alcoholic municipal court alcohol and opiate abuse. London: Gaskell (Royal College of offender. Psychosomatics 1968; 9: 306-10 Psychiatrists), 1993: 19-28 25. Liebson I, Faillace LA. The pharmacological reinforcement of 46. Galanter M. Network therapy for alcohol and drug abuse. Lon- disulfiram – maintenance in chronic alcoholism. NIDA Res don: Guilford Press, 1999: 80 Monogr 1971; 1266-73 47. Brewer C. Managing opiate abuse: learning from other addic- 26. Haynes SN. Contingency management in a municipally admin- tions. J Drug Issues 1988; 18: 679-97 istered Antabuse program for alcoholics. J Behav Ther Exp 48. Brewer C. Combining pharmacological antagonists and behav- Psychiatry 1973; 4: 31-2 ioural psychotherapy in treating addictions: why it is effective 27. Gerrein JR, Rosenberg CM, Manohar V. Disulfiram mainte- but unpopular. Br J Psychiatry 1990; 157: 34-40 nance in outpatient treatment of alcoholism. Arch Gen Psy- 49. Mattson ME, Del Boca FK, Carroll KM, et al. Compliance with chiatry 1973; 28: 798-802 treatment and follow-up protocols in Project MATCH: pre- 28. Azrin NH. Improvements in the community reinforcement ap- dictors and relationship to outcome. Alcohol Clin Exp Res proach to alcoholism. Behav Res Ther 1976; 14: 339-48 1998; 22: 1328-39  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5) Disulfiram in Alcohol Abuse Relapse Prevention 50. Brewer C. Long-term, high-dose disulfiram in the treatment of 58. Poulson EH, Loft S, Andersen JR, et al. Disulfiram therapy – alcohol abuse. Br J Psychiatry 1993; 163: 687-9 adverse drug reactions and interactions. Acta Psychiatr Scand 51. Brewer C. Using disulfiram to maintain controlled drinking: A case report with a 14-year follow-up. Addict Res 1996; 3: 231-5 59. Chick J, Brewer C. National differences in disulfiram prescrib- 52. Gatch MB, Lal H. Pharmacological treatment of alcoholism.
ing. Psychiatric Bull 1999; 23: 335-9 Prog Neuropsychopharmacol Biol Psychiatry 1998; 22: 917-44 60. Brewer C, Hardt F. Preventing disulfiram hepatitis in alcohol 53. Hughes J, Cook C. The efficacy of disulfiram: a review of out- abusers: inappropriate guidelines and the significance of nic-kel allergy. Addict Biol 1999; 4: 303-8 come studies. Addiction 1997; 92 (4): 381-95 61. Larson E, Olincy A, Rummans T, et al. Disulfiram treatment of 54. Liebson IA, Tommasello A, Bigelow GE. A behavioural treat- patients with both alcohol dependence and other psychiatric ment of alcoholic methadone patients. Ann Intern Med 1978; disorders: a review. Alcohol Clin Exp Res 1992; 16: 125-30 62. Gunne L. Politicians and scientists in the combat against drug 55. Litten RZ, Allen J, Fertig J. Pharmacotherapies for alcohol prob- abuse. Drug Alcohol Depend 1990; 25: 241-4 lems: a review of research with focus on developments since1991. Alcohol Clin Exp Res 1996; 20: 859-76 56. Meyers RJ, Smith JE. Clinical guide to alcohol treatment: A community reinforcement approach. New York (NY): Guil- Correspondence and offprints: Dr Colin Brewer, The ford Press, 1995: 57-67 Stapleford Centre, 25a Eccleston Street, London SW1W 57. Branchey L, Davis W, Lee K, et al. Psychiatric complications 9NP, England.
of disulfiram treatment. Am J Psychiatry 1987; 144: 1310-2  Adis International Limited. All rights reserved.
CNS Drugs 2000 Nov; 14 (5)

Source: https://www.robertjmeyersphd.com/pdf/Does%20Disulfiram%20Help%20to%20Prevent%20Relapse%20in%20Alcohol%20Abuse.pdf