By John Grabowski, Howard Rhoades, Ronith Elk, Joy Schmitz, and Daniel Creson

INTRODUCTION

Behavioral Approaches Behavioral interventions for treatment of drug dependence have been developed over the last three decades. They involve systematic manipulation of a broad range of treatment conditions and are important for both understanding and treating features of these disorders.

Their utility resides in the focus on drug-seeking and drugtaking, collection of specific data on individual patterns of drug dependence and abuse, and adapting the elements of treatment to the patients’ needs.

Behavioral approaches also emphasize strategies that actively extend to the patients’ everyday life, thereby assuring that behavioral change will endure after treatment ends.

This is best represented by the work of Higgins and Budney (1993). The terms “behavioral intervention,” “behavior modification,” “behavior therapy,” and “behavioral treatment” often are used interchangeably and describe a variety of therapeutic elements. Each originally emphasized some concepts over others.

While cognitive therapies presume that changing speech and “thinking” changes behavior, behavioral approaches focus on more tangible representations of behavior. All derive from the underlying framework that assumes that altering relationships between behavior and its environmental determinants is critical to change.

Thus, for example, Childress and colleagues (1993) focus on behavior following a model originally proposed for application to drug abuse by Wikler (1948) and O’Brien (1977, 1991).

Changes in responses are presumed to contribute to change in other behaviors such as drug-seeking and drug-taking. Hunt  and Azrin (1973), Stitzer et al. (1985), Bigelow et al. (1984), Higgins and coworkers (1992), Higgins and Budney (1993), and others focused on operant, or instrumental, behavior based on a model explicated by Skinner (1953).

Modification of the behaviors of drug-seeking and drug-taking is presumed to alter physiological and cognitive correlates. Each approach emphasizes specific techniques and is derived from observations and assumptions about determinants of drug abuse.

The unifying strategy is detailed analyses of specific events surrounding drug abuse and application through regimens designed to alter measurable behavior. Within this framework, drugs of abuse are considered to serve as reinforcers, and in that regard they share many of the same characteristics of other events sustaining behavior.

Drug-taking is considered to be an orderly behavior that results from the interaction of fundamental biologic and behavioral processes.

These perspectives have fostered innovative treatment interventions that focus on the interplay of environmental, behavioral, and pharmacological factors and have been the focus of other National Institute on Drug Abuse (NIDA) reviews (e.g., Krasnegor 1979; Thompson and Johanson 1981; Grabowski et al. 1984; Ray 1988). A critical feature of these efforts is the focus on the relationship between drug-taking and its consequences and, in turn, on establishing a similarly persistent relationship between treatment-oriented behavior and its consequences and ensuring that behavioral change endures.

Explicit application of behavioral approaches has made only limited inroads in traditional treatment environments (e.g., Thompson et al. 1984; Schuster and Silverman 1993), despite extensive research and positive findings indicating unique contributions. Unfortunately, to the extent that behavioral interventions have been acknowledged, they often are linked to the elimination of “problem behaviors,” while other therapeutic strategies are proposed to form the basis for developing positive behaviors.

For example, Woody and colleagues (1984) described behavioral interventions in terms of loitering, gun toting, and drug dealing, while psychotherapy was viewed as the intervention of choice for achieving positive goals, The work of Crowley (1984) and, more recently, of Higgins and Budney (1993) points to use of a spectrum of combined therapeutic elements to develop adaptive behaviors. 

All Clinics Have Contingencies

Many clinicians do not refer to their techniques as behavioral interventions, while others simply ignore or eschew behavioral approaches.

However, the very character and framing of therapy dictates that the principles are used, labeled or not. Behavior therapists often refer to “contingencies for reinforcement.”

These specify the relationship between behavior and consequences. Thus, in every clinic, certain behaviors have consequences even if that reaction is punishment or inaction. Typically, contingencies are not recognized or manipulated in a systematic fashion directed at enhancing treatment.

The authors’ thesis is that these often-unspecified contingencies should be made explicit and consistent with treatment goals, then systematically applied. The point of application may range from a single behavior of an individual to an entire treatment system (Thompson et al. 1984).

This chapter illustrates first that common elements of standard treatment settings can be addressed readily from a behavioral perspective rather than being ignored or considered annoyances of clinical systems. The behavioral-environmental features of studies described here were designed, in part, to explicate interventions in the context of standard clinic procedures.

The chapter also describes successive levels of increasingly specific and individualized applications of behavioral approaches. Studies will be summarized with reference to data on clinic function, take-home doses of medications, group contingencies, and, finally, individualized interventions for specific behavioral problems.

A major goal of this chapter is to describe strategies that permit standard drug abuse treatment clinics to operate more effectively (also see Elk et al., in press-a).

GOALS AND IMPLEMENTATION

The goal of implementing clinicwide contingencies is to provide a systemic and systematic foundation for maintenance of treatmentoriented behavior while reducing problems considered endemic and integral to clinics, drug abuse treatment, and the patient population. 

Global fixed contingencies for reinforcement and punishment of a variety of behaviors, both adaptive and maladaptive, can be implemented. Thus, for example, many clinics have provisions for gun toting, loitering, and drug dealing, but they should not be the foremost considerations of treatment. In fact, they can be framed as clinicwide contingencies sustaining adaptive behaviors.

These treatment-oriented behaviors include arriving on time, remaining only for necessary activities, complying with the regimen(s), completing necessary paperwork, providing blood and urine samples when required, and generally using the clinic as a treatment site.

This goal is dependent on the clarity of the contingencies and the consistency of their application and requires no more effort than implementation of traditional systems.

Implementation of manipulated contingencies common to a group of patients similarly requires little more effort than other commonly used formulae. An example is the use of take-home doses of methadone provided contingent on opiate- or cocaine-free urine samples to modify behavior.

A patient who does not use other opiates in one week may receive a specified number of take-home doses in the next week. Conversely, use of opiates during a week when take-home doses of methadone are available produces a requirement for daily visits (no take-home doses) in the subsequent week.

This contingency for all patients receiving methadone can be expected to produce behavioral change in some members of the larger group when systematically applied. Lack of change in behavior of other patients dictates the need for additional or alternative strategies similarly implemented for a subset of patients.

The advantage of this dynamic approach resides in systematic collection of information that then determines modifications of the treatment contingent on specific patient behaviors.

Surprisingly, these techniques are commonly thought to be the unique province of research-oriented clinics. Yet they can be readily applied, and their absence in standard treatment clinics may reduce treatment efficacy. The next level of individualized contingencies in treatment has two stages that may be viewed as fixed contingencies applied to the behavior of individuals or as more refined variable-shaping procedures that are adjusted repeatedly based on one or more individual behavioral patterns.

These may be more labor-intensive interventions, but they are no different from the development of detailed individualized  treatment plans. Based on intake data, individualized elements of treatment can be prescribed.

This orientation provides for sequential implementation of the elements of treatment. For example, elimination of cocaine use in a pregnant, drug-abusing female may be achieved by providing reinforcers for each successive day for which quantitative or semiquantitative urine screens indicate reduced cocaine use.

Compliance with a rigorous medication regimen for a tuberculosis- (TB-) positive male and concurrent reduction in cocaine use may be similarly tailored on an individual basis. The procedures can be implemented sequentially or concurrently. Patients first may be exposed to the general requirements and successively to the specific approaches. Alternatively, all levels may be introduced within days of intake.

IMPLEMENTATION OF CLINICWIDE CONDITIONS OF TREATMENT

Basic “Rules”

In the authors’ clinic, which primarily treats cocaine and opiate dependent patients, the basic contingencies are defined clearly as an integral part of treatment and treatment research (see table 1). Details described in this table permit improved strategies for providing treatment as well as conducting research.

Potential subject-patients participate in an initial telephone screening interview. All procedures are implemented as they might be at any high-quality health care facility, and patients are treated accordingly. Meeting the basic criteria leads to an immediately specified appointment for intake procedures.

The intake process is carefully scheduled and links elements important to inclusion or exclusion. This assures that less-costly elements are carried out first so that false positives are screened out early. This saves patient and staff time as well as money. An initial interview is followed by a general medical evaluation.

The intake process leads to group assignment and medication dispensing for opiate or cocaine dependence at the earliest time, often the same day but always within 24 hours. 

------
TABLE 1 [see source article].
--------

This table lists issues and behaviors that emerge in most drug dependence treatment clinics, although which problems are most frequent is population dependent. Focus on these issues often interferes with service delivery. Generic provisions can be added or eliminated as needed. Positive (“+”) and negative (“-”) consequences must be clearly stated and systematically applied. The goal is specification of positive consequences for which the absence of that consequence is itself unpleasant.

Failure to comply with items 9 and 10 has the consequences of warnings and potential discharge. Some issues, such as discussion of accuracy of laboratory drug screen results, have neither positive nor negative consequences; they are not open for discussion, just as blood pressure readings are medical test results generally accepted and not points for contention.

Examples of Fixed Clinicwide Contingencies for Patients and the Nature of Consequences 1. Regular attendance for continued treatment+/- 2. Maintain appointment time for counseling+/- 3. Maintain appointment time for medication+/- 4. Complete data and information update forms+/- 5. Return of medication bottles+/- 6. Provide urine samples for drug screens as scheduled+/- 7. Arrive and depart in reasonable time (“no loitering”)- 8.

 Maintain clean air (“no smoking”)- 9. Contribute to a physically healthy clinic (“no weapons”)- 10. Support the clinic as the sole vendor (“no drug dealing”)- 11. Responsiveness to chemistry laboratory findings (“no arguing”)o  Patients are provided with medication time options from three 1-hour periods scheduled daily–early morning, midday, or early evening. Similarly, a fixed counseling time is determined in consultation with the patient.

All data and information collection activities are scheduled to reduce inconvenience for the patient. The basic requirements are described in both the intake and consent procedures and in the first counseling session. Urine screens, routinely collected under observation, are described to the patient as essential medical evaluations for treatment and research.

The difficulties of this process have been eased by using a video recording system with tapes reviewed daily by nurses rather than using a live micturition monitor. Finally, there is an inclusive and clearly stated contingency surrounding treatment; it is that 75 percent of all specified data collection points and visits must be met for the patient to be considered to be “in treatment.”

This provides the patient some leeway, assures that there is contact with the provisions of treatment, provides certainty with respect to sufficient data, and assures a clear criterion for dropout. The result of clearly specified contingencies for basic clinic- and treatment-oriented behavior is that only 5 (of over 700) patients have been discharged in the last 4 years of operation due to the serious problems more common in other clinics.

The relatively high rate of compliance independent of individualized treatment contingencies, in part, may be related to a decent and positive environment in which treatment is provided. In brief, clearly labeled, systematically applied contingencies for participation in treatment appear to have a palliative effect in provision of treatment, even in what is construed to be a difficult population.

Large-N Studies With Fixed Group Conditions

The goal of the authors’ primary studies has been to examine specific components of real treatment. The focus of the research is the joint action of behavioral and pharmacological elements.

In a series of studies, the first step was to examine fixed treatment requirements involving visit frequency. The question was whether being required to visit the clinic more frequently (5 days per week) or less frequently (2 days per week) affected outcome. These requirements meet the criterion of contingencies for reinforcement to the extent that a  patient’s behavior determines whether or not he or she will be maintained in treatment.

They are characteristic of standard treatments that typically require a fixed number of visits per week to preclude discharge. Identifying optimal visit frequency is an important issue for cost and effectiveness. Two large studies have addressed the role of take-home dose frequency or visit frequency as a condition of treatment in relation to medication doses (Grabowski et al. 1992a, 19926; Rhoades et al. 1992a).

The importance of medication dose is clear, and medication dose can be expected to interact with other features of treatment. The impetus for examination of take-home dose or visit frequency derives from two sources.

In part, it emanates from findings in behaviorally based studies indicating that contingent take-home doses can be an effective reinforcer (e.g., Bigelow et al. 1984). Beyond potential reinforcing value, this variable was of considerable interest for two pragmatic reasons.

First, it may affect retention, and ultimately success, in treatment independent of manipulating its frequency in a contingent manner. Second, and obviously important, the number of visits per unit time determines the number of patients who can be served by a clinic within available resources. In one ongoing study of primary opiate dependent patients (Rhoades et al. 1992a, two doses of methadone (50 and 80 mg) were examined in relation to two take-home conditions.

Patients received either 2 or 5 take-home doses of methadone each week. Conversely, they were required to visit the clinic either 5 or 2 days per week. Special dispensation was obtained from the Food and Drug Administration and Drug Enforcement Agency to conduct these studies in which takehome requirements other than those provided for by regulation could be implemented; thus, patients were permitted to have take-home doses from the beginning of treatment (Rhoades et al. 1992a).

In a second parallel study with primary cocaine-dependent patients (Grabowski et al. 1992a, submitted), three medication conditions–two fluoxetine doses (20 mg and 40 mg) and placebo–also were examined in relation to two visit conditions–two or five per week. While there was no manipulation of conditions during treatment, the underlying contingencies for reinforcement prevailed as they do in any clinic.

These studies emulate the conditions of traditional clinics, in which failure to comply with the requirements could produce discharge or  dropout status. Conditions were specified clearly and consistently implemented. The major independent variables were medication dose and take-home frequency, while the major dependent measures were retention and illicit drug use. The preliminary interim results for retention with primary opiate dependent patients are presented in figure 1.

Methadone take-home frequency (and hence clinic visit frequency) was clearly a determinant of retention. Patients receiving more take-home doses remained in treatment for longer periods. Opiate-positive urine results did not differ as a function of frequency within groups at the same methadone dose, but they did differ across the two dosage levels.

Opiate-positive urine screens were less frequent, occurring at a rate of about 10 percent, in patients receiving the higher methadone medication dose; the higher dose had the expected effect of greater reductions in illicit opiate use. One of the main results of this study addresses the question raised by

Childress and coworkers (1991) regarding minimum required conditions in methadone maintenance. In this case, within identical counseling, drug screen, and other treatment elements, takehome frequency and dose determined retention.

Minimum required conditions are driven, in part, by patient characteristics and goals, but it is clear that visit frequency can have pronounced effects. The retention results for the study with primary cocaine dependent patients are presented in figure 2 and also indicate the importance of general visit requirements. Clinic visit frequency was a determinant of retention.

Cocaine-positive urine screens did not differ significantly across groups regardless of fluoxetine dose or visit frequency (Grabowski et al., submitted). Framed as either clinic visits or take-home frequency, this variable is a major and important feature of treatment.

It was codified in methadone regulations and also is typically an issue of discussion with respect to cocaine-dependent patients. However, it must be noted with respect to methadone regulations that the requirements are imprecise, are not optimal, and were not substantially data based. Specifically, the low (50 mg) and high (80 mg) doses of methadone combined with high-frequency take-home dose produced the best results with respect to retention. High-frequency take-home doses and, thus, fewer clinic visits also had no untoward or differential effects in cocaine-dependent patients in terms of retention.

Subject Retention by Dose and Take-Home Frequency

FIGURE 1. Patients received low- (2/wk) or high- (5/wk) frequency take-home doses of methadone and visited the clinic either 5 or 2 days/week. Patients in these groups received either 50 mg or 80 mg of methadone. Retention data are presented as a function of these four conditions. KEY: HF=high frequency; LF=low frequency; B=baseline Subject Retention by Dose and Take-Home Frequency

FIGURE 2. Patients received low- (2/wk) or high- (5/wk) frequency take-home doses of methadone and visited the clinic either 5 or 2 days/week. Patients in these groups received either 20 mg or 40 mg of fluoxetine or placebo. Retention data are presented as a function of these six conditions. KEY: HF=high frequency; LF=low frequency; PBO=placebo

There are complexities to the data relevant to treatment. A clinically important collateral finding emerged with the higher dose of methadone. While opiate-positive urine screens were less frequent, cocaine-positive urine screens were more frequent, and this difference was statistically significant.

This does not argue against the use of higher doses of methadone, which contributes to reduced human immunodeficiency virus (HIV) transmission risk, but it does indicate the need for explicit manipulation of contingencies surrounding cocaine use or, at least, careful adjustment of methadone dosing to reach an optimal balance (Grabowski et al. 1993). Placebo was more effective than either dose of fluoxetine within the take-home condition (Grabowski et al., submitted).

An important, but not surprising, collateral finding (Grabowski et al. 1992a) was that benzylecgonine-free urine samples at intake were correlated positively with less drug use during treatment, and, conversely, metabolitecontaining urines at intake were correlated with more drug use during treatment as a function of take-home condition.

Specifically, cocainedependent patients with positive drug screens at intake did less well (i.e., had more cocaine-positive drug screens) than patients who had negative drug screens at intake when required to visit less frequently. This suggests that required level of intervention can be predicted at intake, and this addresses a patient matching issue.

A behavioral perspective stipulates that consideration be given to fine-tuning and individualizing the provisions of treatment, with some patients requiring more visits to the clinic and others fewer. In combination, it is evident that take-home dose frequency is an important factor in retention independent of refined systematic manipulation.

Thus, clinicwide contingencies requiring frequent clinic visits (e.g., 5 or 7 days per week) are arguably less efficacious in terms of generating long-term treatment-oriented behavior.