Addrmor Suhacrorr, Vol. 3, JIP 0 Pergamm Press Ltd 1978
4349 Printed m Great Bntain
PERSONALITY TAILORED COVERT SENSITIZATION OF HEROIN ABUSE LONNIE
R.
University
SNOWDEN* of Oregon
Abstract-Two personality tailored versions of covert sensitization were inserted into ongoing counseling of 42 methadone program participants who continued to abuse heroin. It was expected that treatment built around environmentally focused images would achieve greater success with external locus of control (LOC) persons while treatment built around intraorganismically focused images would produce greater change in internal LOC persons. To further accommodate predicted personality differences, each procedure was introduced with a rationale designed to complement its corresponding LOC orientation. A placebo therapy, involving relaxation training and discussions, controlled for attention and treatment novelty effects. A 2 x 3 analysis of covariance on weekly urine analysis reports for seven post treatment weeks revealed a significant interaction between subject locus of control and treatment type. Findings support the utility of introducing personality adapted covert sensitization into methadone program counseling to counteract persistent heroin abuse.
Heroin abuse has been notoriously intractable to attempts at modification. Addicts relapse at high rates following milieu and expressive psychotherapies (O’Donnell, 1965) for which they are widely regarded as ill suited (e.g. Glasscote, 1971). Partly because of frustration with psychotherapy, replacing heroin with methadone, a long-acting synthetic opiate, has become the treatment of choice. Programs dispensing methadone and assorted social services have been established on a national scale. While methadone programs substantially reduce heroin abuse they do not obliterate it. Clinical and laboratory evidence suggests that some clients retain an appetite for heroin despite high doses of methadone (Martin, Jasinski, Haertzen, Kay, Jones, Mansky.& Carpenter, 1973; Chambers & Taylor, 1972). Thus, one multi-program study found the proportion of heroin using clients to vary between 31% and 50% over successive four month periods (Demaree, 1974). Virtually all methadone programs offer counseling. In a national sample of 23 programs, Sells (1974) found individual counseling, of varying prominence and frequency, in all. Counseling programs have goals covering a spectrum from adjustment to medication through complete resocialization. However ambitious the counseling effort, there is uniform interest in suppressing any heroin use which survives administration of methadone. The conventional response to persistent heroin use is to increase the methadone dose. An effective psychological heroin control technique would present several advantages. These include having clients perceive themselves as the change agent, absence of medical risk and potential applicability to opiate free living if methadone dependence can be phased out. Data from preliminary trials suggests that behaviorally based procedures may produce relatively powerful, reliable curtailment of heroin abuse. Based on a review of treatment literature, which consists heavily of uncontrolled case and small sample reports, Callner (1975) calculated that 43% of the studies conducting follow up found total abstinence. One behavioral counterconditioning technique, covert sensitization (Cautela, 1966), depicts behavioral performance and aversive consequences entirely in imagination. Use of imagery allows personalized representation of contexts and behavior and access to a range of nonarbitrary, personalized aversive stimuli. While evidence from case studies seems encouraging (Steinfeld, 1970; O’Brien & Raynes, 1972; Wisocki, 1973) conclusions remain tentative pending controlled, quantified outcome studies. *Requests
for reprints
should
be sent
to Lonnie
R. Snowden,
97403. 43
University
of Oregon.
Eugene,
Oregon,
44
LONNE
R. SNOWDFN
The present investigations sought to evaluate covert sensitization for modifying heroin abuse among persons maintained on methadone. If successful. the study would recommend a specific counseling procedure for counteracting heroin abuse. A related objective was testing a strategy for matching treatment variations to individual differences. Systematically varying, or tailoring, treatment to individual differences aims at maximizing therapeutic gain by dovetailing treatment and critical features of client personality. Ideally, such research would culminate in optimum treatment variations for client subtypes benefiting less from a standard treatment agent. A departure point for tailoring efforts is identifying client variables predictive of treatment outcome. The necessity of discovering such variables has been cogently argued (Best and Steffy, 1971; Kiesler, 1971). Valid outcome predictors enlighten about therapeutic processes, and form a basis for hypothesizing treatment adjustments to improve the prognosis for poor therapy prospects. Using the locus of control personality variable, Best and Steffy devised treatment variations enhancing persistence of smoking modification gains for internal and external subjects (Best & Steffy, 1975; Best, 1975). They found internal subjects more responsive to a stimulus satiation aversion procedure, while external subjects proved more responsive to an analysis of smoking linked environmental events. The tailoring of treatment to personality proposed in the present study extrapolated the Best and Steffy tailoring to a different addiction using analogous treatment operations. Treatment imagery was varied to emphasize either environmental (social and physical setting, drug paraphernalia) or intraorganismic (physical sensations) contexts of drug abuse. It was hypothesized that an external circumstances focus resembles Best and Steffy’s situational analysis, while pairing of sensations surrounding consumption with unpleasant reactions approximates satiation. It is explicitly recognized that traditional locus of control theorizing does not explain why these treatments should be differentially effective for high and low scorers on the Rotter Scale. The present study attempted to extend a replicated empirical relationship to a different population. To further accommodate expected differences. each procedure was introduced with a justification designed to complement an internal or external orientation. The internal rationale stressed client responsibility in treatment and a willpower interpretation of abandoning drug abuse. The external rationale emphasized therapist responsibility and the avoidance of unmanageable pressures. The context, the problematic behavior, and its therapy were consistently interpreted in accordance with these themes. A third treatment condition controlled for nonspecific effects, and consisted of relaxation training and empathic listening.
METHOD
Subjects were drawn from 2 methadone treatment programs. Both programs offer methadone, medical services, and mandatory eclectic counseling to clients with documented histories of heroin addiction. Personnel in each program helped identify a pool of persons enrolled for more than 2 months and having at least 2 morphine positive findings in their seven previous urine analyses. Urine specimens are routinely collected and analyzed weekly on an irregular schedule. From the pool of qualified clients were recruited 42 persons. These persons volunteered for a “new therapy” which would temporily replace their regular counseling. Subject participation was maintained by appealing to regulations mandating counseling and, ultimately, by paying subjects to complete treatment. The N was maintained at 42. Mean age for the entire sample was 26.3 yr; reported length of addiction averaged 5.2 yr. Subjects had been program clients for an average of 5.6 months. By sex, the sample was 76% male and 24% female. By race, the sample was 83”<; black and 17”/, white.
Personality
tailored
covert
sensitization
45
Factoral representation of LOC (internal vs external) and therapy type (internal focus, external focus. control) formed a 2 x 3 experimental design. Internal and external LOC were defined by a median split of scores from the Rotter (1966) scale. The sample was stratified on number of dirty urines from the seven pre-treatment reports, which ranged from 5-2. Random assignment of seven subjects to each of the 6 cells was conducted from strata. General
treatment
procedure
All subjects participated in 45-60 min sessions conducted by the experimenter, an advanced graduate student in clinical psychology. Time in treatment varied between I* and 3 weeks. Sessions 1 and 2 were virtually identical for experimentals and controls, differing only in the message conveyed as a treatment rationale. During sessions 3-5 treatment groups underwent either of the covert sensitization variations, while controls received further relaxation training and discussion. Cooert sensitization
During Session One the standard client responsibility-will power or therapist responsibility-situational imperatives, treatment rationales were explained to the internal and external treatment groups respectively. For controls, a rationale of similar length attributed heroin abuse to excessive tension. Remaining activities were identical for all subjects and involved introduction to relaxation training, a taped exercise, relaxation rating and assignment of homework. Session 2 was devoted to: (1) playing a second taped relaxation training exercise and securing relaxation ratings, (2) conducting a visual imagery training exercise and (3) conducting a brief standardized interview to identify high temptation cues to form treatment images. Covert sensitization or minimum treatment control commenced during Session 3. For Session 3 the aversive consequence of nausea and infestation with abscesses were each delivered for 4 successive presentations. This was followed by 3 presentations incorporating depression. One presentation of an image suggesting feelings of satisfaction and well-being after refusing dope concluded Session 3 treatment. Covert sensitization was identical in Sessions 4 and 5, except that the ordering of aversive stimuli was varied. The standard Sesseion 4 ordering called for depression (4 presentations), nausea (4 presentations), and abscesses (3 presentations). In Session 5 the ordering was abscesses (4 presentations), depression (4 presentations) and nausea (3 presentations), The final scene in each sequence portrayed positive feelings for turning down an opportunity to use heroin. Image clarity. image aversiveness and depth of relaxation were rated at the conclusion of each session. Ratings were given on a l-5 scale and reflected peak clarity, aversiveness and relaxation. Subjects were instructed to duplicate daily the preceding session’s treatment events as homework. Thus, 12 repetitions, as performed during the session, were assigned. Practice was recorded on a tally sheet, which was reviewed at the next session. Concretely, an internally focused image using abscess aversion included the following excerpt: “Your body is starting to ache the way it does when you want some dope. You feel tense-on edge-like a bundle of nerves. Your back aches; your heart’s pounding hard and fast.” Part of the aversive segment described how “just as you get set to hit your arm that rotten skin breaks loose and slides off. Puss is oozing, seeping out of a big hole in your arm.” An externally focused image began with “You’re sitting in your dining room at the table. Your brother is seated next to you. It’s late afternoon. You can hear your kids playing outside. Your brother reaches into his coat pocket and pulls out the stuff and the works.” On the abscess pairings, this scene culminated in the tissue injury described above.
LONNII: R. SNOWDEN
46
At the conclusion of treatment each subject had received eleven covert aversion trials with each of three noxious events and three trials of covert reward for avoiding heroin use. Subjects were encouraged to conjure the most effective aversive scene when tempted, as defined by specified environmental or intraorganismic events. Thus, subjects were instructed to use the treatment as coping skill. Image recreation was interpreted as self-control under internal treatment and activation of an automatic reaction under external treatment. Attention
control
The content of Sessions 1 and 2 matched closely the content for covert groups. Sessions 3-5 involved continued relaxation training and discussion. menter participated empathically and nondirectively. Dependent
sensitization The experi-
variables.
Ratings of the likelihood of future heroin abuse were obtained at the close of treatment. Subjects gave subjective probabilities “in the forseeable future” and under highly tempting hypothetical circumstances. Subjects also rated treatment usefulness. Beginning the week after the last treatment contact urine analysis results were compiled for 7 weeks. During post-treatment urine collection, counselors at both programs unsystematically chose 2 days in which urine specimens were requested from each client. Urine production was directly observed. One specimen per week was unsystematically discarded while the other was forwarded for testing by the method of thin-layer chromatography. This departed from pre-treatment procedures at one program which had observed pre-treatment urine production only on a spot check basis.
RESULTS
The data were analyzed using analysis of covariance, with pretreatment reports serving as the covariate and post-treatment reports serving at the criterion. Besides providing a more sensitive test, Huck and McLean (1975) have argued that ANACOVA is preferable to change scores or post-treatment only scores in data analysis from pre-post experimental designs. Analysis of covariance on number of morphine positive urine specimens is presented in Table 1. There was no significant main effect for either locus of control (F = 1.97) or treatment type (F = 1.87). The interaction. however, was significant (F = 3.70 P < 0.05). Figure 1 depicts the interaction. Consistent with prediction, the matched LOC treatment focus groups achieved the best results. As a supplement to this demonstration of mean differences, the generality of within-cell change was examined. The dirty urine difference score was calculated for each subject. Cases showing fewer dirty urines post-treatment were called improved, while cases registering an equal or greater number of post-treatment dirty urines were called unimproved. Also, the number of subjects with clean records for all seven post-treatment weeks was tabulated. Proportions of improved and completely urine negative subjects per cell are displayed in Table 2. Assuming no treatment effect, the exact probability of obtaining each cell’s proportion of improved and urine-perfect clients was computed. To obtain a no-covert-sensitization Table
1. Analysis
of covariance
for dirty urines
DF
MS
F
1 2
2.88 2.13
1.87
AxB
2
5.40
3.70’
Error
36
I .46
Source LOC (A) Treatment
(B)
* P c 0.05.
1.91
47
Personality tailored covert sensitization
-
External
l ----*
Internal
z? Internally focused treatment Subject
Externally focused treatment
Control
locus of control by treatment
focus InteractIon
Fig. 1.
base rate for change, records for an arbitrarily chosen 14 consecutive weeks were examined at both participating drug programs. The performance was examined of those clients who met criteria for acceptance into the study but who did not participate (IV = 68). The proportion of clients improving, as previously defined, and reducing morphine positive urines to zero for the second 7 weeks were calculated. The obtained improvement proportion (0.56) and urine perfect proportion (0.16) estimate the base rate of improvement from one 7 week period of the subsequent 7 week period where covert sensitization was not administered. Assuming a binomial distribution about these base rates, the exact probability for improvement and urine perfect proportions observed in each cell was computed. Thus, 100% in the internal LOC-internal treatment focus cell-improved; this outcome, assuming a 56% base rate and a binomial distribution, has probability of < 0.02. The only other cell to approach the conventional 0.05 significance level was the external LOG-external treatment focus cell (P < 0.08).
Table 2. Dirty urines, improvement, and heroin free urine records for each cell
Group Internal Internal Internal External Internal Control External Internal External External External Control
LOC focus LOC focus LOC LOC focus LOC focus LOC
Pre-treatment dirty urines
Post-treatment dirty urines
Fewer post than pre dirty urines (“improved”)
Zero post treatment dirty urines
M 2.86
SD 1.27
M 1.00
SD 1.07
N 7
% 100
N 4
% 57
3.14
0.77
2.14
1.46
5
71
3
43
3.00
1.06
2.71
1.67
3
43
1
14
3.00
1.06
3.00
1.77
3
43
1
14
3.14
0.96
1.86
1.25
6
86
3
43
2.86
0.84
2.57
1.26
4
57
0
0
48
LONNE R. SNOWUEN
Three cells had enough subjects who gave no dirty urines during the post-treatment observation period so that they would occur fewer than 5 times out of 100 assuming a population mean of 0.16 (0.16 estimates the proportion of people showing two or more dirty urines during one 7-week period and zero dirty urines during the subsequent 7 weeks with covert sensitization absent). The cells where P < 0.05 for number of urine perfect people were the internal locus-internally focused treatment cell. the internal locus-externally focused treatment cell, and the external locus-externally focused treatment cell. Thus, on number of subjects showing therapeutic gain, treatment-personality compatibility w-s associated with larger effects. Separate 2 x .) analysis of variance on both subjective probability estimates and the usefulness rating revealed no significant effects. Image clarity, image aversiveness and relaxation ratings, collected to measure key therapeutic processes were analyzed vta analysis of variance. The analyses revealed no significant effects. DISCUSSION
The results support the efficacy of tailored covert sensitization for diminishing heroin abuse among methadone maintained persons. Whether or not the effect exceeds 7 weeks. an improved urine record of this duration can have significant practical consequences (e.g. retention in a methadone treatment). Weekly collection of a urine specimen samples a limited time span and does not tally total heroin abuse. However, this measure is objective and probably more trustworthy than any form of self-report. Moreover, since urinalysis results are used programmatically for planning and decision making, they are intrinsically important. What caused the apparent tailoring effect remains unclear. Best and Steffy (1975) hypothesized differential sensitivity to internal vs external sources of stimulation according to locus of control status. While plausible in accounting for the data from a range of areas. tests capable of direct confirmation are yet to be performed. The contribution of variation in rationale, which was confounded with treatment image variation, remains to be assessed. Internal and external treatment justification emphasized therapist versus client control and responsibility. This distinction parallels the directive and nondirective therapist styles found to interact with LOC in determining treatment effectiveness for group therapy (Abramowitz, Abramowitz. Roback & Jackson, 1974). Determining whether the present findings reflect such an effect awaits future research. An unassessed possible contaminant is treatment unreliability. The issue is whether every instance of therapy matched its treatment description while differing from every therapy instance not of its type. For example, while presented images incorporated only interoceptive or environmental cues developed from contrasting standardized interviews, it remains possible that subjects elaborated images with cues from the excluded class. Such blurring of treatments would produce a discrepancy between the described and effective independent variable and invalidate attributing outcome differences to the experimental manipulation. Future research should evaluate treatment reliability by checking patient perception of images and other treatment operations. A related issue is experimenter bias. Did the experimenter prejudice the results, however unwittingly. in favor of the hypothesis? This could have been accomplished by somehow assisting therapy in locus-focus compatible cells and undermining it in others. The finding of no significant differences in image clarity, image aversiveness, and relaxation partially supports the claim of treatment equivalence. Nevertheless, use of therapists blind to the hypothesis is the only way to exclude an experimenter bias explanation. Subjects in all cells rated their chance of future heroin abuse as quite low (means under either instruction were less than 0.28) and treatment usefulness as quite high (all means above 4 on a 5-point scale). This consistent optimism may be a function of the obvious situational pressures against entertaining or verbalizing any expectation of future drug use. Tailored covert sensitization seems to be feasible and useful as a specialized procedure
Personality
tailored
covert
sensitization
49
for overcoming recalcitrant heroin abuse. It is easily inserted for brief periods into ongoing counseling regardless of counselor philosophy, technique or goals. Client reactions seemed surprisingly positive. The technique was generally judged to be novel and nonthreatening. Clients seemed impressed and rewarded by the discernible impact of relaxation training and image induction. It is not entirely surprising that a brief treatment could undermine a resistant behavior problem like heroin abuse under present conditions. Methadone eliminates the withdrawal-avoidance motive and makes large heroin doses a requisite for achieving pleasurable sensations. Whether tailored covert sensitization would have value for nonmethadone maintained persons is an open question. Limited available evidence suggests that covert sensitization at least as a component of a larger treatment package can alter heroin use. Conceivably tailoring would enhance the treatment effect; confirmation awaits empirical testing. If covert sensitization is therapeutic for methadone abstinent individuals, then its introduction during methadone maintenance would begin skill training under less challenging conditions for coping with abstinence.
Acknowledgemmfs-This article is based on a doctoral dissertation submitted to Wayne State University. The author expresses his appreciation to committee chairer Gerald Rosenbaum and committee members John Teahan, Joel Ager. Mark Goldman, and Donald Marcotte. Gratitude is also extended to Richard Steffy, University of Waterloo. REFERENCES Abramowitz. C. V.. Abramowitz, S. 1.. Roback. H. B. & Jackson, C. Differential effectiveness of directive and nondirective group therapists as a function of client internal-external control. Journal of Consulting and Clinical Psychology, 1974. 42, 8499853. Best, J. A. Tailoring smoking withdrawal procedures to personality and motivational differences, Journal of.Consulting and Clinical Psychology, 1975. 43, 1-8. Best, J. A. & Steffy R. A. Smoking modification procedures tailored to subject characteristics. Behavior Therapy, 1971.2,
17lL191.
Best J. A. & St&y, Canadian
Calmer,
R. A. Smoking
Journal
of Behavioral
D. A. Behavioral
Bulletin,
Cautela.
1975, 82.
treatment
modification Science,
approaches
procedures
for internal
and external
locus of control
clients.
1975, 7, 155-165.
to drug abuse:
A critical
review of the research.
Psylchological
143-l 63.
J. R. Treatment
of compulsive
behavior
by covert
sensitization.
Psychological
Record,
1966.
16,
3341.
Chambers. C. D. & Taylor. W. J. Patterns of “cheating” among methadone maintenance patients. In W. Kemp (Ed.): Drug Abuse: Current Concepts and Research. Springfield. 111.: Charles C. Thomas, 1972. Demaree, R. G. Behavioral measures and related criteria for drug users in the DARP: 1969-1971 admissions. In S. B. Sells, (Ed.): The Efictiueness of Drug Abuse Trentment Vol. 1. Cambridge, Mass.: Ballinger. 1974. Glasscote, R. M. The Treatment of Drug Abuse. Washington D.C.: The Joint Information Service of the American Psychiatric Association and the National Association for Mental Health. 1971. Huck, S. W. & McLean. R. A. Using repeated measures ANOVA to analyze the data from a pretest-posttest design: A potentially confusing task. Psychological Bulletin. 1975. 82, 51 I-518. Kiesler, D. J. Experimental designs in psychotherapy research. In A. Bergin & S. Garfield. (Eds.): Handbook of Psychotherapy and Behavior Change. New York: Wiley, 1971. Martin. W. R., Jasinski, D. R., Haertzen, C. A., Kay, D. C.. Jones, B. E., Mansky. P. A. & Carpenter. R. W. Methadone-a re-evaluation. Archives of General Psychiatry, 1973. 28, 286295. O’Brien, J. S. & Raynes, A. E. Treatment of heroin addiction with behavioral therapy. In W. Keup (Ed.): Drug Abuse: Current Concepts and Research. Springfield. 111.: Charles C. Thomas, 1972. Rotter. J. B. Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs,
1966, 80,
l-28.
Sells, S. B. Research
design. In S. B. Sells. (Ed.): The Bffectiveness of Drug Abuse Trearment Vol. 1. Cambridge, Mass. : Ballinger, 1974. Steinfeld, G. J. The use of covert sensitization with institutionalized narcotic addicts. International Journal of the Addictions, 1970. 5. 225-232. Wisocki. P. A. The successful treatment Behavior Therapy and Experimental
of a heroin Psychiatry.
addict
by covert
1973, 4, 55-61.
conditioning
techniques.
Journal
of