- Email: [email protected]
Predictive factors for outcome of treatment for alcohol problems
Journal of Substance Abuse, 5, 31-44 (1993)
Predictive Factors for Outcome of Treatment for Alcohol Problems Fanny Duckert University of Oslo Institute of Psychology Oslo, Norway
Two samples of problem drinkers were followed up 2 and 4 years after they completed treatment. The first consisted of 72 men and 16 women admitted to a program for alcoholics (the inpatient sample), and the second, of 57 men and 35 women who participated in a program of outpatient treatment (the outpatient sample). At start of treatment, the outpatient sample, was generally characterized by a higher degree of social integration and more moderate alcohol problems than those found in the inpatient sample. These differences were sustained during the part of the follow-up period for which comparative data existed. In both samples it was possible to identify subgroups whose alcohol consumption throughout the observation period did not exceed average consumption in a comparative group of the Norwegian population. The most important predictive factors for alcohol consumption in the inpatient sample were degree of social integration, consumption before start of treatment, and sex. In the outpatient sample the most important factors were level of consumption and relative contribution of heavy drinking to the drinking pattern before start of treatment and the client's own goals as regards to alcohol. In both samples there was a close connection between alcohol consumed, total situation, and individual degree of satisfaction. For both groups, less frequent drinking and reduction of heavy drinking were most important for feeling satisfied with the drinking outcome. The therapeutic implications of the qualitative changes in drinking patterns are discussed. C h o o s i n g o u t c o m e c r i t e r i a is o n e o f t h e d i f f i c u l t p a r t s o f s t u d i e s o n a l c o h o l i s m t r e a t m e n t i n v o l v i n g t h e o r e t i c a l a n d m e t h o d o l o g i c a l q u e s t i o n s . F o r i n s t a n c e , to w h a t e x t e n t s h o u l d a l c o h o l i s m b e d e f i n e d as a u n i t a r y o r m u l t i d i m e n s i o n a l p h e n o m e n o n , a n d s h o u l d t h e effects o f t r e a t m e n t b e e v a l u a t e d solely a g a i n s t t h e c r i t e r i o n o f a b s t i n e n c e ( B a b o r , Dolinsky, R o u n s a v i l l e , & Jaffe, 1988)? I t h a s b e e n q u e s t i o n e d w h e t h e r t o t a l a b s t i n e n c e is s y n o n y m o u s with b e t t e r social f u n c t i o n i n g ( ~ k e r l i n d , H 6 r n q u i s t , E l t o n , & B j u r u l f , 1990; P a t t i s o n , 1977; Polich, A r m o r , & B r a i k e r , 1980). A l t h o u g h it is r e a s o n a b l e to a s s u m e t h a t , t h e h i g h e r t h e c o n s u m p t i o n , t h e m o r e n e g a t i v e t h e o v e r a l l s i t u a t i o n , t h e e x t e n t to w h i c h a b s t i n e n c e is n e c e s s a r y f o r g e n e r a l i m p r o v e m e n t is still n o t s u f f i c i e n t l y clear. I n m o s t s t u d i e s , t h e a s s e s s m e n t o f d r i n k i n g b e h a v i o r is m a d e b y c l a s s i f i c a t i o n s o f t h e s u b j e c t s i n t o a few c a t e g o r i e s . I n t h e s t u d y o f c h a n g e p r o c e s s e s , this
Correspondence and requests for reprints should be sent to Fanny Duckert, University of Oslo, Institute of Psychology, PO Box 1094 Blindern, N-0371 Oslo Norway. 31
32
F. Duckert
procedure may involve a disadvantage in that minor alterations are disregarded (~kerlind et al., 1990). Furthermore, many studies present only partial follow-up data in ostensibly long follow-up studies. Results based on partial intervals may not accurately reflect how subjects actually have fared over the entire period of follow-up (Sobell, Brochu, Sobell, Roy, & Stevens, 1987). Another, often overlooked, element of evaluation is the question of relative changes. For instance, two persons may have had the same level of consumption during one follow-up period. But for one of them this implies no change because he was drinking the same as beforehand, whereas the other one had reduced his cons~amption by 80%. Who is then to be considered most successful? An alternative criterion for substantial change and harm reduction might, for instance, be the demand for a reduction in consumption of more than 50% of pretreatment level. Connected to this is also the question of the individual's own degree of satisfaction with the treatment result. This question has seldom been addressed in evaluation studies, possibly because of a basic mistrust in alcoholics' ability to make adequate assessments of their own situations, but it might be an important criterion for outcome of treatment. If we were to compare the consumption of the participants with the consumption level in the normal population, we might also get an indication of the degree to which their consumption was "normal" or "socially acceptable." In Norway the mean consumption of alcohol, using part of the population, is assumed to be somewhere around 10 L pure ethanol per year. One alternative criterion for outcome could, for instance, be not exceeding 10 L pure alcohol (equivalent to 22 g) per day, which would be about equivalent to the estimated consumption among adult alcohol users in the same area (Skog, 1988) and would imply a low risk for health damage (Royal College of Psychiatrists, 1987). The present work compares data from two different studies of problem drinkers who had been in inpatient and outpatient treatment, respectively. The initial assessment had revealed several important differences between the two samples, as to both drinking patterns and socioeconomic situation. Because the inpatient group had more serious alcohol problems, it was assumed that they more often would choose abstinence as a goal than the outpatient group. It was also expected that they would be more successful in achieving abstinence than reduced consumption, whereas the outpatient group was assumed more often to choose and to achieve reduced consumption. The initial differences also were expected to have consequences for outcome. Thus, the outpatients were expected to have a better outcome in the long run than the inpatients (Duckert, 1988). There were methodological differences between the projects, as for instance the length of the project period and the duration of follow-up periods. Also the projects were partly run in different years: The inpatient study was conducted during the years 1979-1984, and the outpatient study was done during the years 1983-1986. The follow-up procedures differed somewhat. These differences have, of course, restricted the possibility of direct comparisons. However, the similarities in design and methodology gave access to information that might lead to new insights into the nature of and development in different subgroups of problem drinkers.
Predictive Factors for Alcohol Treatment Outcome
33
The present work focuses on prognostic factors for consumption in the two groups and on the relationship between drinking pattern and the participants' own degree of satisfaction. The relationship between alcohol consumption and overall situation is considered, and the two samples are compared to find out whether the same or similar processes occurred in both. Finally, possible therapeutic implications of some of the findings are discussed. METHODS
Subjects The sample comprised 88 persons (72 men and 16 women) from an inpatient alcohol program (the inpatient sample), and 92 persons (57 men and 35 women) from an outpatient clinic (the outpatient sample). All subjects resided in the Oslo area. At the inpatient program the stay of treatment usually lasted 3 - 5 weeks. The treatment was founded on social learning theory, regarding alcohol abuse as a learned and modifiable habit. Relapse-prevention techniques, assertion training, self-monitoring, and social and marital training skills were parts of the treatment (Liberman, King, DeRisi, & McCann, 1975; Marlatt & Gordon, 1985; Miller & Mastria, 1977). Although the official attitude was a preference for total abstinence, there was a certain degree of flexibility with regard to the definition of individual goals. In the outpatient clinic, which was run at the National Institute of Alcohol and Drug Research, a comparison was done between two types of therapeutic intervention: short-term group therapy and individual counseling. Both alternatives were brief treatment choices, based upon social learning theory. All the participants were allowed to make their own choices of treatment goals, whether it was abstinence or reduced consumption, and received some basic information about self-control. The participants were randomized into the two treatment alternatives (Duckert, Amundsen, & Johnsen, 1992). Procedure
Within the year before start of treatment all participants had a medical examination and a clinical interview with special focus on alcohol consumption. Drinking pattern was determined by a modified "time-line follow-back" procedure (Sobell, Maisto, Sobell, & Cooper, 1979) using a calendar to document the number of weeks of abstinence and of moderate and heavy drinking. The participants then had been asked to describe their daily consumption during a representative week of moderate and heavy drinking, respectively. On the basis of this, we could estimate individual alcohol consumption. Both samples had filled in the Severity of Alcohol Dependency Questionnaire (SADQ: Stockwell, Murphy, & Hodgson, 1983), and the outpatient sample had also filled in the Short Michigan Alcoholism Screening Test (SMAST: Seizer, Vinokur, & Van Rooijen, 1975).
34
F. Duckert
After completion of treatment, the inpatient sample was followed up by means of personal interviews every 12 months (the first time, after 6 months), 4 times in all, covering a period of 31/2 years. The outpatient sample was followed up every 6 months (the first time, after 3 months), also 4 times in all, covering a period of 21 months. Additional information was obtained during the follow-up period from the official social security records and the Central Bureau of Statistics. All statistical procedures were done by means of the File Orientated Statistical System (FOSS: Amundsen, 1991). For both samples, multiple regression analyses were carried out, to identify factors that had affected the alcohol consumption, the general situation, and the degree of satisfaction of the participants. Due to incomplete data on some of the participants, the number of subjects was reduced in some of the calculations. The possible effects of dropouts have been discussed more in detail elsewhere (Skog & Duckert, in press). In the analyses, alcohol consumption during the follow-up period, stated degree of satisfaction, and final overall situation were used as dependent variables. Alcohol consumption was defined as estimated average annual consumption during each follow-up period. Due to the uneven distribution of the variable and a clear tendency toward multiplicative effects (Skog & Duckert, in press), alcohol consumption was expressed by means of quasi-logarithmic transformations. In this, abstainers constitute a special problem, because zero is undefined. In the present study, abstainers were arbitrarily given the value zero, and alcohol users were scaled into the interval 1-96. If, as stated initially, alcohol problems are to be considered multidimensional, more extensive criteria for outcome must be used. In this study, we created an index scale, combining different dimensions (codes stated in parentheses): degree of satisfaction with own drinking pattern (0: dissatisfied, 1: partly satisfied, 2: satisfied), number of readmissions during the follow-up period (0: more than 3, 1: 1-3, 2 : 0 admissions), housing conditions (0: no permanent dwelling, 1: permanent dwelling), partnership relationship (0: no stable partnership, 1: stable partnership), employment situation (0: no permanent employment, 1: permanent employment), and mean yearly ethanol consumption throughout the follow-up period (0: > 31 L, 1 : 1 0 - 3 0 L, 2: < 10 L), with the sum: 0-9.
Independent Variables Employment situation, housing situation, partnership relationship, and drinking habits among friends at admission were used as indicators of social integration and social network at start of treatment. Drinking habits were stated partly in terms of quantity of alcohol consumed the last year before admission and partly in terms of how much of this consumption took place in the form of heavy drinking. The degree of alcohol dependence was indicated partly by the results of the SADQ and SMAST tests (for the outpatient sample only, due to lack of information on the inpatient sample) and partly by the clients' own conception of degree of dependence (for the inpatient sample only). In addition, the age of onset of alcohol problems (before the age of 25) and the age of the first contact with
Predictive Factors for Alcohol Treatment Outcome
35
treatment (before the age of 30) were used as indications of early onset of serious alcohol problems (Cloninger, Bohman, & Sigvardsson, 1981; NordstrCm, 1987). Motivating factors were expressed in terms of the clients' stated treatment goals in relation to alcohol. Previous treatment experience was in.dicated by whether the client had had previous contact with alcoholism treatment and by the number of previous alcohol-related admissions. RESULTS Results from the follow-up period (Tables 1 and 2) demonstrated that the initial differences between the inpatient and outpatient samples were sustained. At the end of the observation period, the socioeconomic situation of the outpatients was still generally more positive than that of the inpatient group. During the part of the follow-up period for which comparative data were available, fewer of the outpatient clients had renewed contact with the system of treatment, mortality had been lower than in the inpatient sample, and a larger percentage had paid employment. Alcohol consumption also was still more moderate among the outpatient clients. But in the outpatient sample, our results indicated that, except for a slight short-term effect in favor of the group therapy alternative, small longterm differences in the results were found between the two forms of treatment (Duckert, Johnsen, Amundsen, StrOmme, & M6rland, 1992). In both samples, however, there were large individual differences in final outcome, as regards both drinking behavior and general social situation (Skog & Duckert, in press). Tables 3 and 4 show the bivariate correlations between the independent vari-
Table 1. Social Situation, Partner's Relationship to Alcohol and Drinking Goals at A d m i s s i o n and at the Last Follow-up Interview (Percentage o f Group) Inpafients Adm. End
Outpatients Adm. End
Permanent work
73
53
99
92
Permanent housing
78
89
100
100
Married/stable relationship
34
50
56
55
28 48 18 6
21 40 38 22
12 38 38 13
10 42 40 7
44 30 27
8 43 49
25 56 19
26 62 12
Partner's relationship to alcohol
Abstinent Very moderate "Normal" Problem drinker Stated goals for future relationship to alcohol
Abstinence Reduced consumption No change/unspecificgoals
36
F. Duckert
Table 2. Developmentin Drinking Patterns During Follow-up Periods, Drinking Status, and Degree of Satisfaction With Outcome at the Last Follow-up Interview Inpatients
Outpatients
Estimated mean yearly consumption (litres pure ethanol) Last year before admission During the whole follow-up period
49 25
36 19
Element of heavy drinking in the drinking pattern (%) Last year before admission During the whole follow-up period
70 48
39 21
Change in amount consumed during last follow-up period, compared with the year before admission (percentage of group) More than 50% reduction 20-49% reduction 0-19% reduction More than 20% increase
53 13 13 21
51 20 10 18
Division of subjects on different levels of estimated yearly consumption throughout the whole follow-up period (percentage of group) 0-10 liters ethanol 11-30 liters ethanol More than 31 liters
37 34 29
35 47 18
Proportion of subjects who had been in alcohol-related treatment before admission (percentage of group)
71
34
Proportion of subjects who had alcohol-related hospitalizations during the follow-up period (percentage of group)
69
32
Mean number of rehospitalizations
5.6
Degree of satisfaction with present drinking pattern at last follow-up interview (percentage of group) Very satisfied Partly satisfied Not satisfied
T a b l e 3.
6. 7. 8. 9. 10.
47 26 27
25 36 39
Inpatients: Correlation Matrix Between the Independent Variables
1 1. 2. 3. 4. 5.
0.7
Drinking habits of friends Work/housing/family-situation Pretreatment consumption Percent heavy drinking Previous alcohol-related treatment No. previous admissions Treatment goal Sex Early alcohol-related problems Self-rated dependency
2
-.34 .4.__00 - . 2 3 .16 - . 15 .10 -.01 .03 -.17 .2__.66 .29
.04 -.28 -.20 -.13 - . 15 -.26
3
4
5
6
7
8
9
-.20 -.15
,18
.3__44 -.09 -.13 .3o -.23 -.07 .2__66 .2.__44 - . 2 0 -.03 -.13 - . 2 4 .2___55 .20 .09 .3__99 .16 .15
-.19 .03 .10 .14 - . 0 5 .2..A4 - . 0 2
Note. Correlations with value p < .05 are underlined. See Table 5 for codes.
Predictive Factors for Alcohol Treatment Outcome
Table 4.
37
Outpatients: Correlation Matrix Between the Independent Variables 1
1. Drinking habits of friends 2. Work/house/family situation 3. Pretreatment consumption 4. Percent heavy drinking 5. Previous alcohol-related • treatment 6. No. alcohol-related treatments 7. Treatment goal 8. Sex 9. Early onset of drinking problems 11. SADQ 12. SMAST
-.06 .3.__.66 -.11
2
3
4
5
6
7
.09 .07
-.27
-.03
-.13
.05
.09
-.06 -.04 .02
.01 .07 -.19
-.11 -.05 -.22
.25 -.11 -.12
.40 -.05 -.19
-.05 -.02
.05
.13 .02 -.10
.25 .08 -.05
.25 .01 -.12
.12 .18 .40
.40 .25 .41
.48 .27 .51
.03 -.25 -.16
8
9
11
-.25 -.22 -.20
.20 .41
.52
Note. Correlations with value p < .05 are underlined. See Table 5 for codes•
ables. As can be seen, most of the correlations were moderate. For the inpatient sample, the strongest correlations were found between drinking habits among friends and consumption during the year before treatment. For the outpatient sample, the strongest correlation was between the number of previous alcoholrelated admissions to treatment, SMAST, and early onset of alcohol problems. There was also a moderate correlation between SMAST and SADQ. None of the bivariate correlations were so high, however, as to represent serious danger of multicollinearity in the multivariate analyses.
Alcohol Consumption During the Observation Period Table 5 indicates the relationship between the independent variables and
consumption during the observation period. For the inpatient sample, the strongest bivariate correlations were between alcohol consumption and degree of social integration, drinking habits of friends, self-conceived dependence on alcohol, and amount consumed prior to admission. When the different factors were controlled for the effects of each of the others, social integration, sex, and previous level of consumption turned out to be the strongest predictive factors. For the outpatients, the most important predictive factors were previous drinking pattern, both in amount consumed; degree of heavy drinking; and stated goal. On the basis of the predictive power of the different variables, it seemed that, for the inpatient sample, probability of low consumption during the follow-up period was greatest for persons with a high degree of social integration and a low level of consumption before admission and for females. The probability of high consumption was strongest for persons with poor social integration who had been drinking a great deal before treatment and for men. In the case of the outpatient sample, probability of low consumption was great-
38
F. Duckert
Table 5.
E f f e c t o f D i f f e r e n t F a c t o r s o n A l c o h o l C o n s u m p t i o n During Follow-up Period
Inpatients
Outpatients
(n -- 70)
'1. Drinking habits o f friends (0:heavy drinking or no friends, 1:abstinent/moderate friends) 2. Work/housing/family situation (summed index 0-3 based on: permanent employment[ 1], own dwelling[ 1], stable spouse[l] 3. Alcohol consumption last year before treatment (liters ethanol, log. trans.) 4. Percent heavy drinking last year before treatment 5. Previous alcohol-related treatment
(0:no, 1:yes) 6. Number o f alcohol-related admissions before treatment
7. Treatment goal (0:abstinence, 1:reduction, 2:no change) 8. Sex (l:man, 2:woman) 9. Early onset o f alcohol problems (0:no, 1:yes) 10. Self-rated dependency (0:not/slighdy, l:strongly dependent) 11. SADQ (sum score) 12. SMAST (sum score)
(n --- 89)
r
~
r
13
0.37
0.07
0.16
0.02
-0.49
-0.39
0.02
-0.02
0.41
0.25
0.44
0.45
0.19
-0.08
-0.36
-0.31
0.00
-0.05
0.08
-0.05
0.09
0.05
-0.02
0.03
0.22 -0.25
0.13 -0.29
0.19 -0.07
0.20 -0.03
0.26
0.07
0.22
0.08
0.33
0.07 ---
--0.03 0.01
--0.05 0.22
--
--
Note. Correlation coefficients (Pearson r) and standardized regression coefficients (13). Coefficients with significance value p < .05 are underlined. R* = 0.446 (inpatient); R 2 = 0.380 (outpatient).
est f o r t h o s e w h o h a d h a d low alcohol c o n s u m p t i o n b u t a h i g h d e g r e e o f h e a v y d r i n k i n g b e f o r e start o f t r e a t m e n t a n d w h o s e objective was abstinence. T h e probability o f h i g h c o n s u m p t i o n i n c r e a s e d if alcohol c o n s u m p t i o n h a d b e e n h i g h d u r i n g t h e y e a r p r e c e d i n g start o f t r e a t m e n t a n d if the p e r s o n h a d t h o u g h t to c o n t i n u e as b e f o r e o r h a d only v a g u e t r e a t m e n t goals. Sex s e e m e d to be o f little i m p o r t a n c e . For b o t h samples, the d e g r e e o f d e p e n d e n c e , w h e t h e r m e a s u r e d in t e r m s o f o w n e x p e r i e n c e o r by S A D Q , s e e m e d to be o f little p r e d i c t i v e value. P r e v i o u s admissions f o r a l c o h o l - r e l a t e d p r o b l e m s also s e e m e d to be o f little i m p o r t a n c e . Overall Situation T a b l e 6 shows t h e bivariate a n d m u l t i v a r i a t e c o r r e l a t i o n s b e t w e e n o v e r a l l a n d the i n d e p e n d e n t variables. I n t h e o u t p a t i e n t s a m p l e , p r e t r e a t m e n t alcohol c o n s u m p t i o n s e e m e d to be a s t r o n g predictive f a c t o r f o r final overall situation
Predictive Factors for Alcohol Treatment Outcome Table 6.
39
Effect of Different Factors on Overall Situation at End of Follow-up Period
Outpatients
Inpatients (n = 70)
1. Drinking habits of friends 2. Work/housing/family situation 3. Alcohol consumption last year 4. Percent heavy drinking 5. 'Previous alcohol-related treatment 6. Number of previous alcohol-related admissions 7. Treatment goal 8. Sex 9. Early onset of alcohol problems 10. Self-rated dependency 11. SADQ 12. SMAST
(n = 89)
r
13
r"
13
-0.36 0.49 -0.30 -0.28 -0.14 -0.31 -0.18 0.21 -0.27 -0.38 ---
-0.09 0.30 -0.24 -0.10 -0.09 -0.29 -0.11 0.12 --0.08 ---
0.02 0.09 -0.19 0.15 -0.26 0.29 0.05 -0.03 -0.11 --0.I1 -0.21
0.09 0.05 -0.11 0.15 -0.26 0.50 0.02 -0.06 -0.07 --0.04 -0.30
Note. Correlation coefficients (Pearson r) and standardized regression coefficients (13). Coefficients with significance value p < .05 are underlined. R~ = 0.472 (Inpatients). R2 = 0.338 (Outpatients). See table 5 for codes.
situation. A n o t h e r d e t e r m i n i n g factor in both samples was the n u m b e r o f p r e a d mission alcohol-related t r e a t m e n t experiences, but with opposite effects in the two samples, that is, a negative effect in the inpatient sample a n d a positive effect in the outpatients. Social situation had an i n d e p e n d e n t effect in the i n p a t i e n t material, b u t not in the outpatient. I n the i n p a t i e n t sample, t h e r e were bivariate correlations b e t w e e n overall situation, d r i n k i n g habits o f friends, previous heavy drinking, a n d self-conceived d e p e n d e n c e . But these b e c o m e nonsignificant w h e n controlled for the effects o f the o t h e r factors. I n the outpatient sample, S A D Q was o f equally little value in p r e d i c t i n g the overall situation a n d predicting alcohol c o n s u m p t i o n , whereas S M A S T was o f predictive value. I n the inpatient sample, a positive overall situation at the e n d o f the o b s e r v a t i o n p e r i o d was c o n n e c t e d to low alcohol c o n s u m p t i o n b e f o r e admission a n d few previous alcohol-related admissions b e f o r e the stay as inpatient. T h e probability o f a negative overall situation was greatest for those with a high level o f cons u m p t i o n d u r i n g the follow-up period a n d a history o f m a n y alcohol-related admissions to t r e a t m e n t b e f o r e their stay at the inpatient p r o g r a m . I n the case o f the outpatient sample, the probability o f a positive overall situation at the e n d o f the follow-up p e r i o d was c o n n e c t e d to a low S M A S T score, no previous contact with t r e a t m e n t facilities, o r m a n y previous alcohol-related admissions to t r e a t m e n t . Previous alcohol c o n s u m p t i o n was not a s t r o n g predictor. Sex was a p p a r e n t l y o f little i m p o r t a n c e . T h e chance o f a negative o u t c o m e was greatest f o r those who were socially well i n t e g r a t e d at start o f t r e a t m e n t , h a d a high S M A S T score, a n d h a d previous ( t h o u g h few) alcohol-related admissions to t r e a t m e n t b e f o r e e n t e r i n g the study.
40
F. Duckert
Table 7. Effect of Different Factors on Degree of Satisfaction Among Participants at End of Observation Period
Inpatients (n = 76)
Alcohol consumption during follow-up Percent of drinking that is heavy Percent reduction in consumption during last follow-up period Number of alcohol-related readmissions during follow-up period Drinking less often than before Better withstanding of drinking pressure Less harmful consequences of drinking episodes
Outpatients (n -- 75) r
r
[3
-0.46 -0.60
0.04 -0.35
-0.26 -0.31
-0.14 -0.24
0.04
0.06
0.40
0.21
-0.48 0.40 0.42 0.48
-0.17 0.22 0.22 0.09
0.05 0.43 0.26 0.42
0.10 0.23 0.03 0.22
Note. Correlation coefficients (Pearson r) and standard regression coefficients ([3). Coefficients with value p < .05 are underlined. R~ = 0.489 (inpatient); R~ = 0.318 (outpatient). D e g r e e o f Satisfaction
In the inpatient sample, the most satisfied ones had a m e a n a n n u a l c o n s u m p tion o f 12 L ethanol, o f which 27% was in the f o r m o f heavy drinking, a n d 4.1 readmissions. T h e very dissatisfied ones had a m e a n a n n u a l c o n s u m p t i o n o f 43 L o f which 66% was heavy drinking, a n d a m e a n n u m b e r o f 9.5 readmissions. All these differences were significant. I n the outpatient sample, the most satisfied ones had a m e a n a n n u a l c o n s u m p tion o f 13 L ethanol, o f which 13% was heavy drinking, with a m e a n o f 0.04 readmissions. T h e most dissatisfied ones h a d a m e a n annual c o n s u m p t i o n o f 27 L ethanol, o f which 28% was in the f o r m o f heavy drinking, but with no r e a d m i s sions. A p a r t f r o m the n u m b e r o f readmissions, the differences were significant. Table 7 indicates that for the inpatient sample, the strongest i m p a c t on d e g r e e o f satisfaction c a m e f r o m r e d u c t i o n in heavy d r i n k i n g a n d in d r i n k i n g less o f t e n t h a n before. For the outpatient sample, reductions in heavy drinking, d r i n k i n g less often, a n d less h a r m f u l consequences o f d r i n k i n g episodes were the strongest predictors o f d e g r e e o f satisfaction. For b o t h groups, readmissions were not i m p o r t a n t in p r e d i c t i n g the individual d e g r e e o f satisfaction. T h e r e was s o m e tendency for the persons having e x p e r i e n c e d a high n u m b e r o f readmissions to be m o r e dissatisfied in b o t h samples, b u t having e x p e r i e n c e d o n e readmission did not lead to a lower d e g r e e o f satisfaction with the o u t c o m e . Factors that also were o f little i m p o r t a n c e in b o t h samples were sex, d e g r e e o f dependency, early onset o f p r o b l e m s , possible previous alcohol-related t r e a t m e n t , a n d socioeconomic aspects. DISCUSSION T h e results indicate that, in the p r e s e n t s a m p l e o f clients, t h e r e was a strong relationship between alcohol c o n s u m p t i o n a n d f u t u r e overall situation. T h e i n t e r v e n i n g processes s e e m e d to vary, however, for the two samples.
Predictive Factors for Alcohol Treatment Outcome
41
In the inpatient sample, there seemed to be a more direct correlation between external factors such as degree of social integration and quantity consumed, whereas such factors did not seem to play such an important role in the outpatient sample. For this sample the most important factors seemed to be the nature of their previous drinking patterns and the persons' own goals for future alcohol consumption. It seemed as though the outpatients had greater ability than the inpatients to adjust the amount of alcohol they drank to conform with their own expressed goals. Also, they experienced to a greater degree a reduction in harmful consequences of drinking episodes. Among the inpatients there seemed to be less harmony between what they stated they wanted and what they actually achieved. Similarly, there seemed to be a stronger correlation between alcohol consumption and resulting overall situation in the inpatient sample than in the outpatient sample. This may indicate that alcohol was a more dominating feature in the lives of the inpatients, with a closer link between consumption and social consequences. For the inpatient sample, the frequency of readmissions during the observation period was high, with the largest percentage from the high-consumption group. In the case of the outpatient sample, on the other hand, there were few alcohol-related readmissions during the observation period, and there was little difference in this respect between the different consumption groups, in spite of a slight tendency for an increasing number of readmissions with increasing consumption. In both samples, whether readmissions had taken place or not seemed to have little influence on degree of satisfaction with the drinking outcome, especially in the case of the outpatient participants. This may indicate that, for the subjects themselves, readmissions did not necessarily imply a bad outcome. In both samples, several of the participants reported that, during certain periods, they had been using readmissions as a device for preventing and reducing the extent of heavy drinking. The result of this was that they had been able to reduce their consumption over time. They were satisfied with the final outcome and looked upon the readmissions as a positive change, compared with their old situation, in which they had been more passive and helpless in handling drinking. It was somewhat surprizing that previous contact with treatment facilities before admission had such divergent effects in the two samples. Common to both groups was the fact that previous treatment had a negative effect on the future overall situation (even if this effect was only slight in the inpatient sample). The effect of many pretreatment admissions was different, however, in the two samples. In the inpatient sample, many previous admissions was unambiguously a negative, whereas it had a positive effect in the outpatient sample. A closer look at this factor among the outpatient clients shows that, of the 9 persons with 3 or more previous admissions to treatment, 8 were placed in the upper half of the scale for total social functioning (a score of 4.5 or more). This result cannot be explained, but it is reasonable to assume that it may be connected to the method of recruitment to the project, which clearly described it as a new and different kind of therapeutic project. The majority of the participants had had no previous treatment experience, whereas a small group had had extensive treatment contacts and a relatively low degree of social integration at
42
F. Duckert
start of treatment. It was this latter group who managed especially well throughout the project period. It is possible that these had been "wrongly placed" or "wrongly treated" before or that they had reached "saturation point" with other forms of treatment and were, therefore, looking for, and particularly receptive to, something different. Our data indicate the existence of different subgroups of problem drinkers, with different drinking patterns and different social situations. These differences did have prognostic implications, because the changes over time differed for the two subgroups. Our findings are in agreement with existing literature (Babor et al., 1988; Institute of Medicine, 1990). Outcome Criteria
At admittance, 25% of the outpatient sample and 44% of the inpatient sample had stated a goal of abstinence. During the different follow-up periods, the actual rate of abstinence varied between 6% and 16% in the outpatient group and between 6% and 12% in the inpatient. Very few reported to have been consistently abstinent throughout the whole observation period. Thus, using abstinence as the only criterion for success would not be very useful. However, in both samples the majority significantly reduced their consumption. Another criterion for substantial change and harm reduction might be, as mentioned earlier, a reduction in consumption of more than 50% of pretreatment level. In the two samples, that would have given success rates of 52% in outpatients and 53% in inpatients. During the follow-up period, more than one-third of both the inpatient sample (37%) and the outpatient sample (35%) had a mean annual consumption of less than 10 L ethanol, and 34% and 47%, respectively, had been consuming between 10 and 30 L. Eighteen percent of the outpatient sample, compared to 29% of the inpatient sample, had been drinking more than 30 L. In both samples, the high-consumption group was distinguished by a higher level of consumption before start of treatment. However, the relative reduction in consumption after treatment was greatest for the low- and medium-consumption groups. Our data suggest that there was a strong relationship between drinking and overall situation in both samples, but especially for the inpatient group. Certain changes in drinking behavior, namely reduction in heavy drinking combined with less frequent drinking, were the strongest predictive factors for both samples. The amount consumed in itself and readmissions did not have a strong independent effect. As to the subjects' subjective experiences of satisfaction, abstinence did not seem to be necessary. Relative reductions in amounts consumed, especially in the form of a reduction in heavy drinking, seemed to be the most important factor for being satisfied with the result in both samples. In this work most of the information used was based upon self-reports. This, of course, raises the question of validity. More detailed discussions of this have been done in other works (Duckert et al., 1992; Skog & Duckert, in press). For present concerns, however, it can be mentioned that we found reasonable agree-
Predictive Factors for Alcohol Treatment Outcome
43
m e n t between self-reports and m e a s u r e d Blood Alcohol C o n c e n t r a t i o n (BAC), a n d the biological markers Gamma-Glutamyl T r a n s p e p t i d a s e (GGT), and Average Volume p e r Erythrocyte (MCV). We also f o u n d high a g r e e m e n t between self-reports and official social security records, as to, for instance, t r e a t m e n t admissions (Duckert, Johnsen, et al., 1992).
Therapeutic Implications In both samples, as far as t r e a t m e n t could be assumed to have had an effect on d r i n k i n g behavior, it seemed to a greater d e g r e e to have reached the m o d e r a t e d r i n k i n g part o f the d r i n k i n g r e p e r t o i r e o f the participants. In a way, it is natural that therapies with a rational, cognitive-oriented approach, including strong elements o f educational, self-appraisal, and self-monitoring techniques, mainly will reach the most conscious and rational parts o f the drinking pattern. It may be speculated that the heavy-drinking part is m o r e closely connected with d r i n k i n g urges (or "craving" in the standard literature) established by the principles o f classical conditioning (Hodgson, Stockwell, & Rankin, 1979; Stockwell, Rankin, & Taylor, 1982). H o d g s o n and his colleagues have defined craving as a c o m p l e x system o f interrelated responses involving subjective, physiological, behavioral, and biochemical c o m p o n e n t s that are sometimes only partially coupled. Craving in this orientation is viewed as an obsessive-compulsive disord e r that should be treated in the same way as o t h e r obsessive-compulsive disorders. R e c o m m e n d e d t r e a t m e n t m e t h o d s would be based on systematic, p r o l o n g e d e x p o s u r e to the cues that elicit craving, together with the p r e v e n t i o n o f c o n s u m p t i o n o f alcohol. Unfortunately, we have only a few case studies (Blakey & Baker, 1980; H o d g s o n & Rankin, 1982) r e p o r t i n g the use o f such techniques in treatment. Although these techniques remain to be properly evaluated, they involve some quite interesting elements, omitted in most therapeutic orientations.
REFERENCES ~kerlind, I., H6rnquist, J.O., Elton, M., & Bjurulf, P. (1990). Overall functioning and criteria of progress in rehabilitation of alcohol abusers: Longitudinal analyses of changes. Alcoholism: Clinical and Experimental Research, 14(6), 856-862. Amundsen, A. (1991). Manual for users of FOSS. Oslo: National Institute of Alcohol and Drug Research. Babor, T.F., Dolinsky,Z., Rounsaville, B., &Jaffe, J. (1988). Unitary versus multidimensional models of alcoholism treatment outcome: An empirical study.Journal of Studies on Alcohol, 49(2), 167177. Blake),, R., & Baker, R. (1980). An exposure approach to alcohol abuse. Behaviour Research and Therapy, 18, 319-325. Cloninger, C.R., Bohman, M., & Sigvardsson, S. (1981). Inheritance of alcohol abuse: Cross-fostering analysis of adopted men. Archives of General Psychiatry, 38, 861-868. Duckert, F. (1988). Recruitment to alcohol treatment: A comparison between male and female problem drinkers recruited to treatment in two different ways. British Journal of Addiction, 83, 285-293. Duckert, F. (in press). Alcoholism treatment: Changing or maintaining old habits? A 4-year prospec-
44
F. Ducker/
tive study of men and women admitted to a treatment unit for alcoholics. Scandinavian Journal of Social Medicine. Duckert, E, Amundsen, A., & Johnsen, J. (1992). What happens to drinking after therapeutic intervention? An experimental study of changes in drinking patterns after two out-patient, brief treatment alternatives. British Journal of Addiction, 87, 1457-1467. D,uckert, F., Johnsen, J., Amundsen, A., Str0mme, J., & M6rland, J. (1992). Co-variation between biological markers and self-reported alcohol consumption. Alcohol and Alcoholism, 27, 545555. Hodgson, RJ., & Rankin, HJ. (1982). Cue exposure and relapse prevention. In W.M. Hay & P.E. Nathan, (Eds.), Clinical case studies in the behaviorial treatment of alcoholism. New York: Plenum. Hodgson, RJ., Stockwell, T.R., & Rankin, H.J. (1979). Can alcohol reduce tension? BehaviourResearch and Therapy, 17, 459-466. Institute of Medicine (U.S). (1990). Broadening the base of treatmentfor alcoholproblems. Washington DC: National Academy Press. Liberman, R.P., King, L.W., DeRisi, W.J., & McCann, M. (i 975). Personal effectiveness: Guiding people to assert thewtselvesand improve their social skiU~. Champaign, IL: Research Press. Marlatt, G.A., & Gordon, J.R. (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. New York: Guilford Press. Miller, P.M., & Mastria, M.A. (1977). Alternatives to i~lcoholabuse. Champaign, IL: Research Press. NordstrCm, G. (1988). Successful outcome in alcoholism: A prospective long-term follow-up study. Unpublished doctoral dissertation, University of Lurid, Sweden. Pattison, E.M. (1977). Outcome of Treatment. In E.M. Pattison, M.B. Sobeil, & L.C. Sobell (Eds.), Emerging concepts of alcohol dependence (pp. 257-260). New York: Springer. Polich, J.M., Armor, D.J., & Braiker, H.B. (1980). The course of alcoholism: Four years after treatment. Santa Monica, CA: Rand Corp. The Royal College of Psychiatrists (1986). Alcohol ourfavourite drug. London: Tavistock. Seizer, M.L., Vinokur, A., & Van Rooijen, L. (1975). A self-administered Short Michigan Alcoholism Screening Test (SMAST).Journal of Studies on Alcohol, 36, 117-126. Skog, O.J. (1988). Alcohol related problems in Norway: Patterns and trends. In O.J. Skog & R. Waahlberg (Eds), Alcohol and drugs: The Norwegian experience (47-70). Oslo: ICAA. Skog, O.J., & Duckert, F. (in press). The development of alcoholics' and heavy drinkers' consumption: A longitudinal study. Journal of Studies on Alcohol. Sobell, L., Maisto, S.A., Sobell, M.B., & Cooper, A.M. (1979). Reliability of alcohol abusers' selfreports of drinking behaviour. Behaviour Research and Therapy, 17, 157-160. Sobell, M.B., Brochu, S., SobeU, L., Roy, J., & Stevens, J.A. (1987). Alcohol treatment outcome evaluation methodology: State of the art 1980-1984. Addictive Behaviors, 12(2), 113-128. Stockwell, T.R., Rankin, H., & Taylor, C. (1982). Alcohol dependence, beliefs and the priming effect. Behaviour Research and Therapy, 20, 513-522. Stockwell, T., Murphy, D., & Hodgson, R. (1983). The severity of alcohol dependence questionnaire: Its use, reliability and validity. BritishJourual of Addiction, 78, 145-155.
Predictive Factors for Outcome of Treatment for Alcohol Problems Fanny Duckert University of Oslo Institute of Psychology Oslo, Norway
Two samples of problem drinkers were followed up 2 and 4 years after they completed treatment. The first consisted of 72 men and 16 women admitted to a program for alcoholics (the inpatient sample), and the second, of 57 men and 35 women who participated in a program of outpatient treatment (the outpatient sample). At start of treatment, the outpatient sample, was generally characterized by a higher degree of social integration and more moderate alcohol problems than those found in the inpatient sample. These differences were sustained during the part of the follow-up period for which comparative data existed. In both samples it was possible to identify subgroups whose alcohol consumption throughout the observation period did not exceed average consumption in a comparative group of the Norwegian population. The most important predictive factors for alcohol consumption in the inpatient sample were degree of social integration, consumption before start of treatment, and sex. In the outpatient sample the most important factors were level of consumption and relative contribution of heavy drinking to the drinking pattern before start of treatment and the client's own goals as regards to alcohol. In both samples there was a close connection between alcohol consumed, total situation, and individual degree of satisfaction. For both groups, less frequent drinking and reduction of heavy drinking were most important for feeling satisfied with the drinking outcome. The therapeutic implications of the qualitative changes in drinking patterns are discussed. C h o o s i n g o u t c o m e c r i t e r i a is o n e o f t h e d i f f i c u l t p a r t s o f s t u d i e s o n a l c o h o l i s m t r e a t m e n t i n v o l v i n g t h e o r e t i c a l a n d m e t h o d o l o g i c a l q u e s t i o n s . F o r i n s t a n c e , to w h a t e x t e n t s h o u l d a l c o h o l i s m b e d e f i n e d as a u n i t a r y o r m u l t i d i m e n s i o n a l p h e n o m e n o n , a n d s h o u l d t h e effects o f t r e a t m e n t b e e v a l u a t e d solely a g a i n s t t h e c r i t e r i o n o f a b s t i n e n c e ( B a b o r , Dolinsky, R o u n s a v i l l e , & Jaffe, 1988)? I t h a s b e e n q u e s t i o n e d w h e t h e r t o t a l a b s t i n e n c e is s y n o n y m o u s with b e t t e r social f u n c t i o n i n g ( ~ k e r l i n d , H 6 r n q u i s t , E l t o n , & B j u r u l f , 1990; P a t t i s o n , 1977; Polich, A r m o r , & B r a i k e r , 1980). A l t h o u g h it is r e a s o n a b l e to a s s u m e t h a t , t h e h i g h e r t h e c o n s u m p t i o n , t h e m o r e n e g a t i v e t h e o v e r a l l s i t u a t i o n , t h e e x t e n t to w h i c h a b s t i n e n c e is n e c e s s a r y f o r g e n e r a l i m p r o v e m e n t is still n o t s u f f i c i e n t l y clear. I n m o s t s t u d i e s , t h e a s s e s s m e n t o f d r i n k i n g b e h a v i o r is m a d e b y c l a s s i f i c a t i o n s o f t h e s u b j e c t s i n t o a few c a t e g o r i e s . I n t h e s t u d y o f c h a n g e p r o c e s s e s , this
Correspondence and requests for reprints should be sent to Fanny Duckert, University of Oslo, Institute of Psychology, PO Box 1094 Blindern, N-0371 Oslo Norway. 31
32
F. Duckert
procedure may involve a disadvantage in that minor alterations are disregarded (~kerlind et al., 1990). Furthermore, many studies present only partial follow-up data in ostensibly long follow-up studies. Results based on partial intervals may not accurately reflect how subjects actually have fared over the entire period of follow-up (Sobell, Brochu, Sobell, Roy, & Stevens, 1987). Another, often overlooked, element of evaluation is the question of relative changes. For instance, two persons may have had the same level of consumption during one follow-up period. But for one of them this implies no change because he was drinking the same as beforehand, whereas the other one had reduced his cons~amption by 80%. Who is then to be considered most successful? An alternative criterion for substantial change and harm reduction might, for instance, be the demand for a reduction in consumption of more than 50% of pretreatment level. Connected to this is also the question of the individual's own degree of satisfaction with the treatment result. This question has seldom been addressed in evaluation studies, possibly because of a basic mistrust in alcoholics' ability to make adequate assessments of their own situations, but it might be an important criterion for outcome of treatment. If we were to compare the consumption of the participants with the consumption level in the normal population, we might also get an indication of the degree to which their consumption was "normal" or "socially acceptable." In Norway the mean consumption of alcohol, using part of the population, is assumed to be somewhere around 10 L pure ethanol per year. One alternative criterion for outcome could, for instance, be not exceeding 10 L pure alcohol (equivalent to 22 g) per day, which would be about equivalent to the estimated consumption among adult alcohol users in the same area (Skog, 1988) and would imply a low risk for health damage (Royal College of Psychiatrists, 1987). The present work compares data from two different studies of problem drinkers who had been in inpatient and outpatient treatment, respectively. The initial assessment had revealed several important differences between the two samples, as to both drinking patterns and socioeconomic situation. Because the inpatient group had more serious alcohol problems, it was assumed that they more often would choose abstinence as a goal than the outpatient group. It was also expected that they would be more successful in achieving abstinence than reduced consumption, whereas the outpatient group was assumed more often to choose and to achieve reduced consumption. The initial differences also were expected to have consequences for outcome. Thus, the outpatients were expected to have a better outcome in the long run than the inpatients (Duckert, 1988). There were methodological differences between the projects, as for instance the length of the project period and the duration of follow-up periods. Also the projects were partly run in different years: The inpatient study was conducted during the years 1979-1984, and the outpatient study was done during the years 1983-1986. The follow-up procedures differed somewhat. These differences have, of course, restricted the possibility of direct comparisons. However, the similarities in design and methodology gave access to information that might lead to new insights into the nature of and development in different subgroups of problem drinkers.
Predictive Factors for Alcohol Treatment Outcome
33
The present work focuses on prognostic factors for consumption in the two groups and on the relationship between drinking pattern and the participants' own degree of satisfaction. The relationship between alcohol consumption and overall situation is considered, and the two samples are compared to find out whether the same or similar processes occurred in both. Finally, possible therapeutic implications of some of the findings are discussed. METHODS
Subjects The sample comprised 88 persons (72 men and 16 women) from an inpatient alcohol program (the inpatient sample), and 92 persons (57 men and 35 women) from an outpatient clinic (the outpatient sample). All subjects resided in the Oslo area. At the inpatient program the stay of treatment usually lasted 3 - 5 weeks. The treatment was founded on social learning theory, regarding alcohol abuse as a learned and modifiable habit. Relapse-prevention techniques, assertion training, self-monitoring, and social and marital training skills were parts of the treatment (Liberman, King, DeRisi, & McCann, 1975; Marlatt & Gordon, 1985; Miller & Mastria, 1977). Although the official attitude was a preference for total abstinence, there was a certain degree of flexibility with regard to the definition of individual goals. In the outpatient clinic, which was run at the National Institute of Alcohol and Drug Research, a comparison was done between two types of therapeutic intervention: short-term group therapy and individual counseling. Both alternatives were brief treatment choices, based upon social learning theory. All the participants were allowed to make their own choices of treatment goals, whether it was abstinence or reduced consumption, and received some basic information about self-control. The participants were randomized into the two treatment alternatives (Duckert, Amundsen, & Johnsen, 1992). Procedure
Within the year before start of treatment all participants had a medical examination and a clinical interview with special focus on alcohol consumption. Drinking pattern was determined by a modified "time-line follow-back" procedure (Sobell, Maisto, Sobell, & Cooper, 1979) using a calendar to document the number of weeks of abstinence and of moderate and heavy drinking. The participants then had been asked to describe their daily consumption during a representative week of moderate and heavy drinking, respectively. On the basis of this, we could estimate individual alcohol consumption. Both samples had filled in the Severity of Alcohol Dependency Questionnaire (SADQ: Stockwell, Murphy, & Hodgson, 1983), and the outpatient sample had also filled in the Short Michigan Alcoholism Screening Test (SMAST: Seizer, Vinokur, & Van Rooijen, 1975).
34
F. Duckert
After completion of treatment, the inpatient sample was followed up by means of personal interviews every 12 months (the first time, after 6 months), 4 times in all, covering a period of 31/2 years. The outpatient sample was followed up every 6 months (the first time, after 3 months), also 4 times in all, covering a period of 21 months. Additional information was obtained during the follow-up period from the official social security records and the Central Bureau of Statistics. All statistical procedures were done by means of the File Orientated Statistical System (FOSS: Amundsen, 1991). For both samples, multiple regression analyses were carried out, to identify factors that had affected the alcohol consumption, the general situation, and the degree of satisfaction of the participants. Due to incomplete data on some of the participants, the number of subjects was reduced in some of the calculations. The possible effects of dropouts have been discussed more in detail elsewhere (Skog & Duckert, in press). In the analyses, alcohol consumption during the follow-up period, stated degree of satisfaction, and final overall situation were used as dependent variables. Alcohol consumption was defined as estimated average annual consumption during each follow-up period. Due to the uneven distribution of the variable and a clear tendency toward multiplicative effects (Skog & Duckert, in press), alcohol consumption was expressed by means of quasi-logarithmic transformations. In this, abstainers constitute a special problem, because zero is undefined. In the present study, abstainers were arbitrarily given the value zero, and alcohol users were scaled into the interval 1-96. If, as stated initially, alcohol problems are to be considered multidimensional, more extensive criteria for outcome must be used. In this study, we created an index scale, combining different dimensions (codes stated in parentheses): degree of satisfaction with own drinking pattern (0: dissatisfied, 1: partly satisfied, 2: satisfied), number of readmissions during the follow-up period (0: more than 3, 1: 1-3, 2 : 0 admissions), housing conditions (0: no permanent dwelling, 1: permanent dwelling), partnership relationship (0: no stable partnership, 1: stable partnership), employment situation (0: no permanent employment, 1: permanent employment), and mean yearly ethanol consumption throughout the follow-up period (0: > 31 L, 1 : 1 0 - 3 0 L, 2: < 10 L), with the sum: 0-9.
Independent Variables Employment situation, housing situation, partnership relationship, and drinking habits among friends at admission were used as indicators of social integration and social network at start of treatment. Drinking habits were stated partly in terms of quantity of alcohol consumed the last year before admission and partly in terms of how much of this consumption took place in the form of heavy drinking. The degree of alcohol dependence was indicated partly by the results of the SADQ and SMAST tests (for the outpatient sample only, due to lack of information on the inpatient sample) and partly by the clients' own conception of degree of dependence (for the inpatient sample only). In addition, the age of onset of alcohol problems (before the age of 25) and the age of the first contact with
Predictive Factors for Alcohol Treatment Outcome
35
treatment (before the age of 30) were used as indications of early onset of serious alcohol problems (Cloninger, Bohman, & Sigvardsson, 1981; NordstrCm, 1987). Motivating factors were expressed in terms of the clients' stated treatment goals in relation to alcohol. Previous treatment experience was in.dicated by whether the client had had previous contact with alcoholism treatment and by the number of previous alcohol-related admissions. RESULTS Results from the follow-up period (Tables 1 and 2) demonstrated that the initial differences between the inpatient and outpatient samples were sustained. At the end of the observation period, the socioeconomic situation of the outpatients was still generally more positive than that of the inpatient group. During the part of the follow-up period for which comparative data were available, fewer of the outpatient clients had renewed contact with the system of treatment, mortality had been lower than in the inpatient sample, and a larger percentage had paid employment. Alcohol consumption also was still more moderate among the outpatient clients. But in the outpatient sample, our results indicated that, except for a slight short-term effect in favor of the group therapy alternative, small longterm differences in the results were found between the two forms of treatment (Duckert, Johnsen, Amundsen, StrOmme, & M6rland, 1992). In both samples, however, there were large individual differences in final outcome, as regards both drinking behavior and general social situation (Skog & Duckert, in press). Tables 3 and 4 show the bivariate correlations between the independent vari-
Table 1. Social Situation, Partner's Relationship to Alcohol and Drinking Goals at A d m i s s i o n and at the Last Follow-up Interview (Percentage o f Group) Inpafients Adm. End
Outpatients Adm. End
Permanent work
73
53
99
92
Permanent housing
78
89
100
100
Married/stable relationship
34
50
56
55
28 48 18 6
21 40 38 22
12 38 38 13
10 42 40 7
44 30 27
8 43 49
25 56 19
26 62 12
Partner's relationship to alcohol
Abstinent Very moderate "Normal" Problem drinker Stated goals for future relationship to alcohol
Abstinence Reduced consumption No change/unspecificgoals
36
F. Duckert
Table 2. Developmentin Drinking Patterns During Follow-up Periods, Drinking Status, and Degree of Satisfaction With Outcome at the Last Follow-up Interview Inpatients
Outpatients
Estimated mean yearly consumption (litres pure ethanol) Last year before admission During the whole follow-up period
49 25
36 19
Element of heavy drinking in the drinking pattern (%) Last year before admission During the whole follow-up period
70 48
39 21
Change in amount consumed during last follow-up period, compared with the year before admission (percentage of group) More than 50% reduction 20-49% reduction 0-19% reduction More than 20% increase
53 13 13 21
51 20 10 18
Division of subjects on different levels of estimated yearly consumption throughout the whole follow-up period (percentage of group) 0-10 liters ethanol 11-30 liters ethanol More than 31 liters
37 34 29
35 47 18
Proportion of subjects who had been in alcohol-related treatment before admission (percentage of group)
71
34
Proportion of subjects who had alcohol-related hospitalizations during the follow-up period (percentage of group)
69
32
Mean number of rehospitalizations
5.6
Degree of satisfaction with present drinking pattern at last follow-up interview (percentage of group) Very satisfied Partly satisfied Not satisfied
T a b l e 3.
6. 7. 8. 9. 10.
47 26 27
25 36 39
Inpatients: Correlation Matrix Between the Independent Variables
1 1. 2. 3. 4. 5.
0.7
Drinking habits of friends Work/housing/family-situation Pretreatment consumption Percent heavy drinking Previous alcohol-related treatment No. previous admissions Treatment goal Sex Early alcohol-related problems Self-rated dependency
2
-.34 .4.__00 - . 2 3 .16 - . 15 .10 -.01 .03 -.17 .2__.66 .29
.04 -.28 -.20 -.13 - . 15 -.26
3
4
5
6
7
8
9
-.20 -.15
,18
.3__44 -.09 -.13 .3o -.23 -.07 .2__66 .2.__44 - . 2 0 -.03 -.13 - . 2 4 .2___55 .20 .09 .3__99 .16 .15
-.19 .03 .10 .14 - . 0 5 .2..A4 - . 0 2
Note. Correlations with value p < .05 are underlined. See Table 5 for codes.
Predictive Factors for Alcohol Treatment Outcome
Table 4.
37
Outpatients: Correlation Matrix Between the Independent Variables 1
1. Drinking habits of friends 2. Work/house/family situation 3. Pretreatment consumption 4. Percent heavy drinking 5. Previous alcohol-related • treatment 6. No. alcohol-related treatments 7. Treatment goal 8. Sex 9. Early onset of drinking problems 11. SADQ 12. SMAST
-.06 .3.__.66 -.11
2
3
4
5
6
7
.09 .07
-.27
-.03
-.13
.05
.09
-.06 -.04 .02
.01 .07 -.19
-.11 -.05 -.22
.25 -.11 -.12
.40 -.05 -.19
-.05 -.02
.05
.13 .02 -.10
.25 .08 -.05
.25 .01 -.12
.12 .18 .40
.40 .25 .41
.48 .27 .51
.03 -.25 -.16
8
9
11
-.25 -.22 -.20
.20 .41
.52
Note. Correlations with value p < .05 are underlined. See Table 5 for codes•
ables. As can be seen, most of the correlations were moderate. For the inpatient sample, the strongest correlations were found between drinking habits among friends and consumption during the year before treatment. For the outpatient sample, the strongest correlation was between the number of previous alcoholrelated admissions to treatment, SMAST, and early onset of alcohol problems. There was also a moderate correlation between SMAST and SADQ. None of the bivariate correlations were so high, however, as to represent serious danger of multicollinearity in the multivariate analyses.
Alcohol Consumption During the Observation Period Table 5 indicates the relationship between the independent variables and
consumption during the observation period. For the inpatient sample, the strongest bivariate correlations were between alcohol consumption and degree of social integration, drinking habits of friends, self-conceived dependence on alcohol, and amount consumed prior to admission. When the different factors were controlled for the effects of each of the others, social integration, sex, and previous level of consumption turned out to be the strongest predictive factors. For the outpatients, the most important predictive factors were previous drinking pattern, both in amount consumed; degree of heavy drinking; and stated goal. On the basis of the predictive power of the different variables, it seemed that, for the inpatient sample, probability of low consumption during the follow-up period was greatest for persons with a high degree of social integration and a low level of consumption before admission and for females. The probability of high consumption was strongest for persons with poor social integration who had been drinking a great deal before treatment and for men. In the case of the outpatient sample, probability of low consumption was great-
38
F. Duckert
Table 5.
E f f e c t o f D i f f e r e n t F a c t o r s o n A l c o h o l C o n s u m p t i o n During Follow-up Period
Inpatients
Outpatients
(n -- 70)
'1. Drinking habits o f friends (0:heavy drinking or no friends, 1:abstinent/moderate friends) 2. Work/housing/family situation (summed index 0-3 based on: permanent employment[ 1], own dwelling[ 1], stable spouse[l] 3. Alcohol consumption last year before treatment (liters ethanol, log. trans.) 4. Percent heavy drinking last year before treatment 5. Previous alcohol-related treatment
(0:no, 1:yes) 6. Number o f alcohol-related admissions before treatment
7. Treatment goal (0:abstinence, 1:reduction, 2:no change) 8. Sex (l:man, 2:woman) 9. Early onset o f alcohol problems (0:no, 1:yes) 10. Self-rated dependency (0:not/slighdy, l:strongly dependent) 11. SADQ (sum score) 12. SMAST (sum score)
(n --- 89)
r
~
r
13
0.37
0.07
0.16
0.02
-0.49
-0.39
0.02
-0.02
0.41
0.25
0.44
0.45
0.19
-0.08
-0.36
-0.31
0.00
-0.05
0.08
-0.05
0.09
0.05
-0.02
0.03
0.22 -0.25
0.13 -0.29
0.19 -0.07
0.20 -0.03
0.26
0.07
0.22
0.08
0.33
0.07 ---
--0.03 0.01
--0.05 0.22
--
--
Note. Correlation coefficients (Pearson r) and standardized regression coefficients (13). Coefficients with significance value p < .05 are underlined. R* = 0.446 (inpatient); R 2 = 0.380 (outpatient).
est f o r t h o s e w h o h a d h a d low alcohol c o n s u m p t i o n b u t a h i g h d e g r e e o f h e a v y d r i n k i n g b e f o r e start o f t r e a t m e n t a n d w h o s e objective was abstinence. T h e probability o f h i g h c o n s u m p t i o n i n c r e a s e d if alcohol c o n s u m p t i o n h a d b e e n h i g h d u r i n g t h e y e a r p r e c e d i n g start o f t r e a t m e n t a n d if the p e r s o n h a d t h o u g h t to c o n t i n u e as b e f o r e o r h a d only v a g u e t r e a t m e n t goals. Sex s e e m e d to be o f little i m p o r t a n c e . For b o t h samples, the d e g r e e o f d e p e n d e n c e , w h e t h e r m e a s u r e d in t e r m s o f o w n e x p e r i e n c e o r by S A D Q , s e e m e d to be o f little p r e d i c t i v e value. P r e v i o u s admissions f o r a l c o h o l - r e l a t e d p r o b l e m s also s e e m e d to be o f little i m p o r t a n c e . Overall Situation T a b l e 6 shows t h e bivariate a n d m u l t i v a r i a t e c o r r e l a t i o n s b e t w e e n o v e r a l l a n d the i n d e p e n d e n t variables. I n t h e o u t p a t i e n t s a m p l e , p r e t r e a t m e n t alcohol c o n s u m p t i o n s e e m e d to be a s t r o n g predictive f a c t o r f o r final overall situation
Predictive Factors for Alcohol Treatment Outcome Table 6.
39
Effect of Different Factors on Overall Situation at End of Follow-up Period
Outpatients
Inpatients (n = 70)
1. Drinking habits of friends 2. Work/housing/family situation 3. Alcohol consumption last year 4. Percent heavy drinking 5. 'Previous alcohol-related treatment 6. Number of previous alcohol-related admissions 7. Treatment goal 8. Sex 9. Early onset of alcohol problems 10. Self-rated dependency 11. SADQ 12. SMAST
(n = 89)
r
13
r"
13
-0.36 0.49 -0.30 -0.28 -0.14 -0.31 -0.18 0.21 -0.27 -0.38 ---
-0.09 0.30 -0.24 -0.10 -0.09 -0.29 -0.11 0.12 --0.08 ---
0.02 0.09 -0.19 0.15 -0.26 0.29 0.05 -0.03 -0.11 --0.I1 -0.21
0.09 0.05 -0.11 0.15 -0.26 0.50 0.02 -0.06 -0.07 --0.04 -0.30
Note. Correlation coefficients (Pearson r) and standardized regression coefficients (13). Coefficients with significance value p < .05 are underlined. R~ = 0.472 (Inpatients). R2 = 0.338 (Outpatients). See table 5 for codes.
situation. A n o t h e r d e t e r m i n i n g factor in both samples was the n u m b e r o f p r e a d mission alcohol-related t r e a t m e n t experiences, but with opposite effects in the two samples, that is, a negative effect in the inpatient sample a n d a positive effect in the outpatients. Social situation had an i n d e p e n d e n t effect in the i n p a t i e n t material, b u t not in the outpatient. I n the i n p a t i e n t sample, t h e r e were bivariate correlations b e t w e e n overall situation, d r i n k i n g habits o f friends, previous heavy drinking, a n d self-conceived d e p e n d e n c e . But these b e c o m e nonsignificant w h e n controlled for the effects o f the o t h e r factors. I n the outpatient sample, S A D Q was o f equally little value in p r e d i c t i n g the overall situation a n d predicting alcohol c o n s u m p t i o n , whereas S M A S T was o f predictive value. I n the inpatient sample, a positive overall situation at the e n d o f the o b s e r v a t i o n p e r i o d was c o n n e c t e d to low alcohol c o n s u m p t i o n b e f o r e admission a n d few previous alcohol-related admissions b e f o r e the stay as inpatient. T h e probability o f a negative overall situation was greatest for those with a high level o f cons u m p t i o n d u r i n g the follow-up period a n d a history o f m a n y alcohol-related admissions to t r e a t m e n t b e f o r e their stay at the inpatient p r o g r a m . I n the case o f the outpatient sample, the probability o f a positive overall situation at the e n d o f the follow-up p e r i o d was c o n n e c t e d to a low S M A S T score, no previous contact with t r e a t m e n t facilities, o r m a n y previous alcohol-related admissions to t r e a t m e n t . Previous alcohol c o n s u m p t i o n was not a s t r o n g predictor. Sex was a p p a r e n t l y o f little i m p o r t a n c e . T h e chance o f a negative o u t c o m e was greatest f o r those who were socially well i n t e g r a t e d at start o f t r e a t m e n t , h a d a high S M A S T score, a n d h a d previous ( t h o u g h few) alcohol-related admissions to t r e a t m e n t b e f o r e e n t e r i n g the study.
40
F. Duckert
Table 7. Effect of Different Factors on Degree of Satisfaction Among Participants at End of Observation Period
Inpatients (n = 76)
Alcohol consumption during follow-up Percent of drinking that is heavy Percent reduction in consumption during last follow-up period Number of alcohol-related readmissions during follow-up period Drinking less often than before Better withstanding of drinking pressure Less harmful consequences of drinking episodes
Outpatients (n -- 75) r
r
[3
-0.46 -0.60
0.04 -0.35
-0.26 -0.31
-0.14 -0.24
0.04
0.06
0.40
0.21
-0.48 0.40 0.42 0.48
-0.17 0.22 0.22 0.09
0.05 0.43 0.26 0.42
0.10 0.23 0.03 0.22
Note. Correlation coefficients (Pearson r) and standard regression coefficients ([3). Coefficients with value p < .05 are underlined. R~ = 0.489 (inpatient); R~ = 0.318 (outpatient). D e g r e e o f Satisfaction
In the inpatient sample, the most satisfied ones had a m e a n a n n u a l c o n s u m p tion o f 12 L ethanol, o f which 27% was in the f o r m o f heavy drinking, a n d 4.1 readmissions. T h e very dissatisfied ones had a m e a n a n n u a l c o n s u m p t i o n o f 43 L o f which 66% was heavy drinking, a n d a m e a n n u m b e r o f 9.5 readmissions. All these differences were significant. I n the outpatient sample, the most satisfied ones had a m e a n a n n u a l c o n s u m p tion o f 13 L ethanol, o f which 13% was heavy drinking, with a m e a n o f 0.04 readmissions. T h e most dissatisfied ones h a d a m e a n annual c o n s u m p t i o n o f 27 L ethanol, o f which 28% was in the f o r m o f heavy drinking, but with no r e a d m i s sions. A p a r t f r o m the n u m b e r o f readmissions, the differences were significant. Table 7 indicates that for the inpatient sample, the strongest i m p a c t on d e g r e e o f satisfaction c a m e f r o m r e d u c t i o n in heavy d r i n k i n g a n d in d r i n k i n g less o f t e n t h a n before. For the outpatient sample, reductions in heavy drinking, d r i n k i n g less often, a n d less h a r m f u l consequences o f d r i n k i n g episodes were the strongest predictors o f d e g r e e o f satisfaction. For b o t h groups, readmissions were not i m p o r t a n t in p r e d i c t i n g the individual d e g r e e o f satisfaction. T h e r e was s o m e tendency for the persons having e x p e r i e n c e d a high n u m b e r o f readmissions to be m o r e dissatisfied in b o t h samples, b u t having e x p e r i e n c e d o n e readmission did not lead to a lower d e g r e e o f satisfaction with the o u t c o m e . Factors that also were o f little i m p o r t a n c e in b o t h samples were sex, d e g r e e o f dependency, early onset o f p r o b l e m s , possible previous alcohol-related t r e a t m e n t , a n d socioeconomic aspects. DISCUSSION T h e results indicate that, in the p r e s e n t s a m p l e o f clients, t h e r e was a strong relationship between alcohol c o n s u m p t i o n a n d f u t u r e overall situation. T h e i n t e r v e n i n g processes s e e m e d to vary, however, for the two samples.
Predictive Factors for Alcohol Treatment Outcome
41
In the inpatient sample, there seemed to be a more direct correlation between external factors such as degree of social integration and quantity consumed, whereas such factors did not seem to play such an important role in the outpatient sample. For this sample the most important factors seemed to be the nature of their previous drinking patterns and the persons' own goals for future alcohol consumption. It seemed as though the outpatients had greater ability than the inpatients to adjust the amount of alcohol they drank to conform with their own expressed goals. Also, they experienced to a greater degree a reduction in harmful consequences of drinking episodes. Among the inpatients there seemed to be less harmony between what they stated they wanted and what they actually achieved. Similarly, there seemed to be a stronger correlation between alcohol consumption and resulting overall situation in the inpatient sample than in the outpatient sample. This may indicate that alcohol was a more dominating feature in the lives of the inpatients, with a closer link between consumption and social consequences. For the inpatient sample, the frequency of readmissions during the observation period was high, with the largest percentage from the high-consumption group. In the case of the outpatient sample, on the other hand, there were few alcohol-related readmissions during the observation period, and there was little difference in this respect between the different consumption groups, in spite of a slight tendency for an increasing number of readmissions with increasing consumption. In both samples, whether readmissions had taken place or not seemed to have little influence on degree of satisfaction with the drinking outcome, especially in the case of the outpatient participants. This may indicate that, for the subjects themselves, readmissions did not necessarily imply a bad outcome. In both samples, several of the participants reported that, during certain periods, they had been using readmissions as a device for preventing and reducing the extent of heavy drinking. The result of this was that they had been able to reduce their consumption over time. They were satisfied with the final outcome and looked upon the readmissions as a positive change, compared with their old situation, in which they had been more passive and helpless in handling drinking. It was somewhat surprizing that previous contact with treatment facilities before admission had such divergent effects in the two samples. Common to both groups was the fact that previous treatment had a negative effect on the future overall situation (even if this effect was only slight in the inpatient sample). The effect of many pretreatment admissions was different, however, in the two samples. In the inpatient sample, many previous admissions was unambiguously a negative, whereas it had a positive effect in the outpatient sample. A closer look at this factor among the outpatient clients shows that, of the 9 persons with 3 or more previous admissions to treatment, 8 were placed in the upper half of the scale for total social functioning (a score of 4.5 or more). This result cannot be explained, but it is reasonable to assume that it may be connected to the method of recruitment to the project, which clearly described it as a new and different kind of therapeutic project. The majority of the participants had had no previous treatment experience, whereas a small group had had extensive treatment contacts and a relatively low degree of social integration at
42
F. Duckert
start of treatment. It was this latter group who managed especially well throughout the project period. It is possible that these had been "wrongly placed" or "wrongly treated" before or that they had reached "saturation point" with other forms of treatment and were, therefore, looking for, and particularly receptive to, something different. Our data indicate the existence of different subgroups of problem drinkers, with different drinking patterns and different social situations. These differences did have prognostic implications, because the changes over time differed for the two subgroups. Our findings are in agreement with existing literature (Babor et al., 1988; Institute of Medicine, 1990). Outcome Criteria
At admittance, 25% of the outpatient sample and 44% of the inpatient sample had stated a goal of abstinence. During the different follow-up periods, the actual rate of abstinence varied between 6% and 16% in the outpatient group and between 6% and 12% in the inpatient. Very few reported to have been consistently abstinent throughout the whole observation period. Thus, using abstinence as the only criterion for success would not be very useful. However, in both samples the majority significantly reduced their consumption. Another criterion for substantial change and harm reduction might be, as mentioned earlier, a reduction in consumption of more than 50% of pretreatment level. In the two samples, that would have given success rates of 52% in outpatients and 53% in inpatients. During the follow-up period, more than one-third of both the inpatient sample (37%) and the outpatient sample (35%) had a mean annual consumption of less than 10 L ethanol, and 34% and 47%, respectively, had been consuming between 10 and 30 L. Eighteen percent of the outpatient sample, compared to 29% of the inpatient sample, had been drinking more than 30 L. In both samples, the high-consumption group was distinguished by a higher level of consumption before start of treatment. However, the relative reduction in consumption after treatment was greatest for the low- and medium-consumption groups. Our data suggest that there was a strong relationship between drinking and overall situation in both samples, but especially for the inpatient group. Certain changes in drinking behavior, namely reduction in heavy drinking combined with less frequent drinking, were the strongest predictive factors for both samples. The amount consumed in itself and readmissions did not have a strong independent effect. As to the subjects' subjective experiences of satisfaction, abstinence did not seem to be necessary. Relative reductions in amounts consumed, especially in the form of a reduction in heavy drinking, seemed to be the most important factor for being satisfied with the result in both samples. In this work most of the information used was based upon self-reports. This, of course, raises the question of validity. More detailed discussions of this have been done in other works (Duckert et al., 1992; Skog & Duckert, in press). For present concerns, however, it can be mentioned that we found reasonable agree-
Predictive Factors for Alcohol Treatment Outcome
43
m e n t between self-reports and m e a s u r e d Blood Alcohol C o n c e n t r a t i o n (BAC), a n d the biological markers Gamma-Glutamyl T r a n s p e p t i d a s e (GGT), and Average Volume p e r Erythrocyte (MCV). We also f o u n d high a g r e e m e n t between self-reports and official social security records, as to, for instance, t r e a t m e n t admissions (Duckert, Johnsen, et al., 1992).
Therapeutic Implications In both samples, as far as t r e a t m e n t could be assumed to have had an effect on d r i n k i n g behavior, it seemed to a greater d e g r e e to have reached the m o d e r a t e d r i n k i n g part o f the d r i n k i n g r e p e r t o i r e o f the participants. In a way, it is natural that therapies with a rational, cognitive-oriented approach, including strong elements o f educational, self-appraisal, and self-monitoring techniques, mainly will reach the most conscious and rational parts o f the drinking pattern. It may be speculated that the heavy-drinking part is m o r e closely connected with d r i n k i n g urges (or "craving" in the standard literature) established by the principles o f classical conditioning (Hodgson, Stockwell, & Rankin, 1979; Stockwell, Rankin, & Taylor, 1982). H o d g s o n and his colleagues have defined craving as a c o m p l e x system o f interrelated responses involving subjective, physiological, behavioral, and biochemical c o m p o n e n t s that are sometimes only partially coupled. Craving in this orientation is viewed as an obsessive-compulsive disord e r that should be treated in the same way as o t h e r obsessive-compulsive disorders. R e c o m m e n d e d t r e a t m e n t m e t h o d s would be based on systematic, p r o l o n g e d e x p o s u r e to the cues that elicit craving, together with the p r e v e n t i o n o f c o n s u m p t i o n o f alcohol. Unfortunately, we have only a few case studies (Blakey & Baker, 1980; H o d g s o n & Rankin, 1982) r e p o r t i n g the use o f such techniques in treatment. Although these techniques remain to be properly evaluated, they involve some quite interesting elements, omitted in most therapeutic orientations.
REFERENCES ~kerlind, I., H6rnquist, J.O., Elton, M., & Bjurulf, P. (1990). Overall functioning and criteria of progress in rehabilitation of alcohol abusers: Longitudinal analyses of changes. Alcoholism: Clinical and Experimental Research, 14(6), 856-862. Amundsen, A. (1991). Manual for users of FOSS. Oslo: National Institute of Alcohol and Drug Research. Babor, T.F., Dolinsky,Z., Rounsaville, B., &Jaffe, J. (1988). Unitary versus multidimensional models of alcoholism treatment outcome: An empirical study.Journal of Studies on Alcohol, 49(2), 167177. Blake),, R., & Baker, R. (1980). An exposure approach to alcohol abuse. Behaviour Research and Therapy, 18, 319-325. Cloninger, C.R., Bohman, M., & Sigvardsson, S. (1981). Inheritance of alcohol abuse: Cross-fostering analysis of adopted men. Archives of General Psychiatry, 38, 861-868. Duckert, F. (1988). Recruitment to alcohol treatment: A comparison between male and female problem drinkers recruited to treatment in two different ways. British Journal of Addiction, 83, 285-293. Duckert, F. (in press). Alcoholism treatment: Changing or maintaining old habits? A 4-year prospec-
44
F. Ducker/
tive study of men and women admitted to a treatment unit for alcoholics. Scandinavian Journal of Social Medicine. Duckert, E, Amundsen, A., & Johnsen, J. (1992). What happens to drinking after therapeutic intervention? An experimental study of changes in drinking patterns after two out-patient, brief treatment alternatives. British Journal of Addiction, 87, 1457-1467. D,uckert, F., Johnsen, J., Amundsen, A., Str0mme, J., & M6rland, J. (1992). Co-variation between biological markers and self-reported alcohol consumption. Alcohol and Alcoholism, 27, 545555. Hodgson, RJ., & Rankin, HJ. (1982). Cue exposure and relapse prevention. In W.M. Hay & P.E. Nathan, (Eds.), Clinical case studies in the behaviorial treatment of alcoholism. New York: Plenum. Hodgson, RJ., Stockwell, T.R., & Rankin, H.J. (1979). Can alcohol reduce tension? BehaviourResearch and Therapy, 17, 459-466. Institute of Medicine (U.S). (1990). Broadening the base of treatmentfor alcoholproblems. Washington DC: National Academy Press. Liberman, R.P., King, L.W., DeRisi, W.J., & McCann, M. (i 975). Personal effectiveness: Guiding people to assert thewtselvesand improve their social skiU~. Champaign, IL: Research Press. Marlatt, G.A., & Gordon, J.R. (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. New York: Guilford Press. Miller, P.M., & Mastria, M.A. (1977). Alternatives to i~lcoholabuse. Champaign, IL: Research Press. NordstrCm, G. (1988). Successful outcome in alcoholism: A prospective long-term follow-up study. Unpublished doctoral dissertation, University of Lurid, Sweden. Pattison, E.M. (1977). Outcome of Treatment. In E.M. Pattison, M.B. Sobeil, & L.C. Sobell (Eds.), Emerging concepts of alcohol dependence (pp. 257-260). New York: Springer. Polich, J.M., Armor, D.J., & Braiker, H.B. (1980). The course of alcoholism: Four years after treatment. Santa Monica, CA: Rand Corp. The Royal College of Psychiatrists (1986). Alcohol ourfavourite drug. London: Tavistock. Seizer, M.L., Vinokur, A., & Van Rooijen, L. (1975). A self-administered Short Michigan Alcoholism Screening Test (SMAST).Journal of Studies on Alcohol, 36, 117-126. Skog, O.J. (1988). Alcohol related problems in Norway: Patterns and trends. In O.J. Skog & R. Waahlberg (Eds), Alcohol and drugs: The Norwegian experience (47-70). Oslo: ICAA. Skog, O.J., & Duckert, F. (in press). The development of alcoholics' and heavy drinkers' consumption: A longitudinal study. Journal of Studies on Alcohol. Sobell, L., Maisto, S.A., Sobell, M.B., & Cooper, A.M. (1979). Reliability of alcohol abusers' selfreports of drinking behaviour. Behaviour Research and Therapy, 17, 157-160. Sobell, M.B., Brochu, S., SobeU, L., Roy, J., & Stevens, J.A. (1987). Alcohol treatment outcome evaluation methodology: State of the art 1980-1984. Addictive Behaviors, 12(2), 113-128. Stockwell, T.R., Rankin, H., & Taylor, C. (1982). Alcohol dependence, beliefs and the priming effect. Behaviour Research and Therapy, 20, 513-522. Stockwell, T., Murphy, D., & Hodgson, R. (1983). The severity of alcohol dependence questionnaire: Its use, reliability and validity. BritishJourual of Addiction, 78, 145-155.