Best Practices : Substance Abuse Treatment and Rehabilitation

1. The Effectiveness of Specific Treatment Approaches

While most people who experience substance abuse problems do not receive help, there is good evidence that people exposed to some types of treatment subsequently reduce their use of psychoactive substances and show improvement in other life areas. In general, treatment outcomes are improved when appropriate treatments are also provided for significant life problems (communications problems, lack of assertiveness, unemployment). This section will review in some detail the research available on the effectiveness of a number of the more prevalent and commonly studied treatment modalities.

There have been many hundreds of studies of treatment for people with alcohol and other drug problems and a complete review of all published studies is beyond the scope of this report. However, such a review is not considered necessary given the quality and comprehensiveness of recent published reviews. Much of what follows draws on several of these reviews and especially on two complementary reviews of studies of treatment for alcohol problems by Holder et al. (1991), Miller et al. (1995) and a review of many of the same studies by Finney and Monahan (1996). The present document also draws on the review of treatment studies for dependence on drugs other than alcohol by the National Institute of Drug Abuse (1996), and a review of treatment studies involving adolescents by Smart (1993).

Excellent reviews of studies of treatment for people with alcohol or drug problems by Eliany and Rush (1992), Landry (1995) and the Correctional Service of Canada (CSC)¹ (1996) also provided a wealth of ideas and some useful summary statements. Where appropriate, recent reviews of specific modalities and issues will be cited under the appropriate topic headings.

The Holder et al. (1991) and Miller et al. (1995) reviews of alcoholism treatment encompassed the same 219 studies. Only studies that compared at least one specific type of treatment with some alternative type were included. Studies were excluded if they did not use a proper procedure to equate treatment, if they did not use control groups (randomization or case-control), or if they did not use at least one outcome measure of drinking or drinking problems. Studies without control or comparison groups were not considered because they cannot show how the study subjects in treatment might have fared without treatment or when treated in other ways. However, many of the studies that were reviewed had other shortcomings that limit their interpretation. These shortcomings will be noted at various points in what follows and at the end of Section 7.

Both the Holder et al. and Miller et al. reviews include tables that summarize the cumulative evidence with respect to specific treatment methods and use the same effectiveness index. Weighted Evidence Index scores were computed by counting the number of controlled studies for specific interventions and subtracting the number showing no effects from the number showing positive effects on at least one alcohol-related measure. An extra point was then added for every positive finding greater than 2. Thus, a modality for which there were four studies showing positive results and three with negative results would receive a score of 4-3+2=3. This "box-score" approach has the advantage of simplicity, but it fails to take account of the magnitude of effects and is not sensitive to the number of statistical tests performed.² The approach also takes no account of the quality of the studies considered, the types of comparisons made or the characteristics of those involved in different studies.

Parts of two summary tables from the Holder et al. review are reproduced in Table 1. The results differ slightly from those in the Miller et al. review due to differences in the classification of interventions. Further details of these results and notes on more recent studies of alcoholism treatment will be considered below under a variety of headings.

Finney and Monahan (1996) reviewed many of the same studies as Holder et al. (1991) and developed an alternative index to assess treatment effectiveness. Unlike the simple "box score" index used by Holder et al., this index took account of the percentage of studies with positive results, the number of statistical tests performed and the type of comparisons. This report will note where these two reviews draw similar and varied conclusions. Table 2 provides a comparison of the rankings arrived at in the two reviews.

There are far fewer controlled trials of treatment for dependence on drugs other than alcohol and there would be little gained by constructing tables comparable to those in the Holder et al. and Miller et al. reviews of treatment for alcohol problems. Conclusions from the NIDA (1996), Smart (1993), Landry (1995) and CSC (1996) reviews will therefore be considered with respect to specific types of intervention.

The remainder of this section is structured using the classification scheme for treatment modalities proposed by Holder et al. with the addition of a few other modalities (Table 3).

Table 1. Treatment modalities reviewed by Holder et al., (1991); Number of studies,
studies with positive results and Weighted Evidence Index

	na	+b	WEInc
Good evidence of effect (+6 or higher)
Social skills training	10	10	+18
Self-control training	17	12	+17
Brief motivational counselling	9	8	+13
Marital therapy, behavioural	7	7	+12
Community reinforcement approach	4	4	+6
Stress management training	10	6	+6
Fair evidence of effect (+2 to +5)
Aversion therapy, covert sensitization	7	4	+3
Behaviour contracting	4	3	+3
Disulfiram, oral	10	5	+3
Psychotropic medication, antidepressants	4	3	+3
Disulfiram, implants	5	3	+2
Indeterminate evidence of effect (-1 to +1)
Marital therapy, other (non-behavioural)	3	1	+1
Psychotropic medication, lithium	6	3	+1
Cognitive therapy	7	3	0
Hypnosis	4	2	0
Insufficient evidence (fewer than 3 studies)
Acupuncture	1	1	+1
Calcium carbimide	1	1	+1
Residential milieu, Minnesota model	1	1	+1
Residential milieu, halfway house	1	0	-1
Alcoholics Anonymous	2	0	-2
Aversion therapy apnoea	2	0	-2
Psychotropic medication, antipsychotic	2	0	-2
No evidence of effect (-2 or lower)
Aversion therapy, electrical	15	5	-2
Aversion therapy, chemical (nausea)	5	1	-3
Confrontational interventions	4	0	-4
Psychotherapy (individual)	8	2	-4
Psychotropic medication, psychedelic	8	2	-4
Videotape self-confrontation	4	0	-4
Educational lectures/films	9	2	-5
Psychotropic medication, anti-anxiety	10	2	-6
Counselling, general	9	1	-7
Metronidazole	10	1	-8
Group psychotherapy	13	2	-9
Residential milieu treatment	14	1	-12

Table 2. Rankings by effectiveness indices of 24 treatment modalities reviewed
by Holder et al. (1991) and by Finney and Monahan (1996)

Holder et al. Indexa	Modality	Modality	Finney and Monahan Indexb
18	Social skills training	Community reinforcement approach	59
17	Self-control training	Social skills training	37
13	Brief motivational counselling	Marital therapy, behavioural	36
12	Marital therapy, behavioural	Disulfiram, implants	34
6	Community reinforcement approach	Marital therapy, other (non-behavioural)	21
6	Stress management training	Stress management or other relaxation training	12
3	Disulfiram, oral	Aversion therapy, chemical (nausea)	3
3	Aversion therapy, covert sensitization	Psychotropic medication, antidepressants	2
3	Psychotropic medication,antidepressants	Psychotropic medication, lithium	-2
2	Disulfiram, implants	Brief motivational counselling	-4
1	Marital therapy, other (non-behavioural)	Aversion therapy, covert sensitization	-5
0	Cognitive therapy	Aversion therapy, electrical	-5
0	Hypnosis	Self-control training	-7
1	Psychotropic medication, lithium	Cognitive therapy	-8
-2	Aversion therapy, electrical	Educational films/lectures	-11
-3	Aversion therapy, chemical (nausea)	Group Therapy	-13
-4	Confrontational interventions	Psychotropic medication, LSD	-15
-4	Psychotropic medication, psychedelic	Psychotropic medication, anti-anxiety	-17
-5	Educational lectures/films	Metronidazole	-21
-6	Psychotropic medication, anti-anxiety	Disulfiram, oral	-27
-7	Counselling, general	Residential milieu	-27
-8	Metronidazole	Confrontational interventions	-31
-9	Group psychotherapy	Counselling, general	-32
-12	Residential milieu treatment	Hypnosis	-37

It is important to note that most of the studies considered have not used complete abstinence as the sole or even principal measure of success. Rather, they have used multiple and usually continuous outcome measures (e.g., percentage of drinking days, amount consumed per drinking occasion). This reflects a view of substance abuse as a chronic relapsing condition for which goals of either improvement or "cure" may be appropriate. As with other chronic conditions (e.g., asthma, obesity or late onset of diabetes), it may be unrealistic to expect that substance abuse will be completely or permanently eliminated following a single intervention. However, significant improvements may follow appropriate interventions and these will be detected with the use of appropriate measures.

The increasing use of multiple and continuous outcome measures also reflects an emerging consensus about what to expect from treatment. Some reasonable goals have been proposed by McLellan et al. (1996), including the expectation that treatment should be of benefit not only to the person treated, but to those who are affected by this person, such as family members, health care providers and insurers, employers, and those who work in the correctional system. Outcomes of interest should therefore relate to alcohol and drug use, health and social functioning, use of health services and threats to public safety. To be considered successful, treatment should result in some reductions in at least one of these domains - preferably without a corresponding increase in other domains. In some cases, changes in risk behaviours and in harms associated with drug use (e.g., infections due to needle sharing or accidents due to impaired driving) may be a more realistic goal than complete abstinence, or even reduced substance use. In other cases, reduced or "controlled" substance use may be a realistic expectation, especially for those who are not heavily dependent.

Outcome measures used in the studies reviewed below varied considerably. This is a regrettable feature of substance abuse treatment outcome research and seriously limits comparisons of the results from different studies.

Table 3. Classification scheme used for the review of specific treatment modalities
according to type

Type of Modality	Specific Modalities
Pharmacotherapies	Antidipsotropic Drug Therapy for Alcohol: - disulfiram, implants, oral - metronidazole - calcium carbimide Anti-craving Drug Therapy for Alcohol: - fluoxetine, zimelidine, citalopram - buspirone, ritansarin - naltrexone* Psychotropic Drug Therapy for Alcohol: - antianxiety drugs - lithium - antipsychotic drugs - antidepressants - psychedelics Treatment for Other Drugs: - buprenorphine, naloxone - bupropion - heroin - methadone - clonidine, naltrexone, LAAM, codeine
Behaviour therapies	Aversion Therapies: - covert sensitization - electrical aversion - nausea induction Other Behaviour Therapies: - behavioural contracting - behavioural relapse prevention - behavioural self- control training - cognitive therapy - community reinforcement approach - cue exposure - marital behavioural therapy - social skills training - stress management - video self-confrontation
Psychotherapeutic approaches	Counselling, general Confrontational Interventions Group Psychotherapy Individual Insight-Oriented Psychotherapy
Other approaches	Brief Motivational Counselling Education Hypnosis Residential Milieu Therapy: - Minnesota model - therapeutic community* Alcoholics Anonymous