|Home | About | Journals | Submit | Contact Us | Français|
This article provides experiential evidence on the transportability of the Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST) screening tool and brief intervention in a mental health clinic. There is very little published information on implementing screening and brief intervention (SBI) in a mental health setting. Moreover, few SBI projects have reported on clinicians’ experiences using the ASSIST. The article documents a successful attempt at implementing the ASSIST and discusses the benefits and challenges of doing SBI in a mental health setting.
The ASSIST has gained renown in the past few years as an interview-administered screening tool for primary care settings. The ASSIST is currently mandated by the Substance Abuse and Mental Health Services Administration for all screening and brief intervention (SBI) grants and promoted by the White House Office of National Drug Control Policy and the World Health Organization (WHO ASSIST Working Group, 2002; World Health Organization, 2008). It is widely assumed that the ASSIST can be used in diverse health service settings by a range of professionals. Little research exists, however, on the transportability of the ASSIST in primary care or other health care venues. Transportability research examines the “movement of efficacious interventions to usual-care settings” (Schoewald & Hoagwood, 2001) and is concerned with the who, what, when, and where of innovation implementation and its effects on clients and systems (Schoenwald & Hoagwood). The purpose of this article is to describe how mental health clinicians implemented the ASSIST in a university counseling center.
The research reported here is based on experiential evidence from clinicians who administered the ASSIST with college students and ongoing documentation of SBI activity. This information formed the basis of a process evaluation that we conducted for the “UCLA Access to Care” project, which was funded by a 3-year Substance Abuse and Mental Health Services Administration (SAMHSA) services grant. Our project was 1 of 12 that implemented SBI services on college campuses. The goals of our process evaluation were to (1) document actual use of the ASSIST, (2) understand how mental health clinicians used the ASSIST and what adaptations they made, and (3) explore the functions of the ASSIST and the meanings that clinicians constructed around doing the ASSIST in a mental health setting. This article is the first to document the implementation of the ASSIST in an ongoing clinical practice.
As a recent report from the WHO’s ASSIST Working Group acknowledged, integration of SBI into primary care practice has proved to be a “difficult business” (Anderson et al., 2004; Heather, 2007). Some of the challenges reported by primary care clinicians include lack of time, lack of training, low self-efficacy in delivering interventions, lack of psychiatric referral sources, attitudes about substance users, and views that substance use counseling is not part of the physician’s role (McAvoy et al., 2001; Neushotz & Fitzpatrick, 2008). Similar barriers have been cited for conducting SBI in trauma centers.
To promote the implementation of SBI, a number of alternative models have been developed that use health educators and other non-medical staff to conduct SBI. One of the main discoveries driving the development of alternative models was that screening does not need to be administered by physicians or nurses (Babor et al., 2007). The Academic Emergency Department Screening, Brief Intervention, and Referral to Treatment (SBIRT) Research Collaborative used research assistants to do the screening in emergency departments. The emergency department programs engaged physicians, nurses, and other hospital staff in delivery of the brief intervention, which is clinically oriented (Academic ED SBIRT Research Collaborative, 2007). As another example, an SBI project for college students organized self-administered screening through the campus health center and used counseling center clinicians to conduct the brief intervention (Martens et al., 2007).
There is also support for self-administration by patients using computers. Dimeff and McNeely (2000) found positive results with a computerized, self-administered screening approach combined with in-person feedback and advice from physicians in a student health center. Moreover, there is growing support for computerized SBI methods that require no person-to-person contact (Hester, Squires, & Delaney, 2005; Walters, Miller & Chiauzzi, 2005). The computerized methods are used outside the context of health services as stand-alone programs. SBI implementation methods are diverse and usually tailored to fit the circumstances of the venue and the target populations.
There appears to be a growing trend to outsource SBI to outside personnel (e.g., dedicated research assistants and health educators; Madras et al., 2008; Sise, Sise, Kelley, Simmons, & Kelso, 2005). The original design of the WHO’s SBI initiative, however, was to integrate SBI into the behavioral repertoires of primary care staff. Integration implied a change in the behavior of primary care staff to adopt this evidence-based practice (Babor & Higgins-Biddle, 2000; Heather, 2007). The goal for the WHO and many researchers worldwide was to increase the “uptake” of screening by primary care providers (Heather, 2007; Roche & Freeman, 2004). In the area of emergency care, there has also been a strong interest in having medical professionals administer some part of the SBI process (DiClemente, 2005).
Most of the current implementation work on SBI, however, has moved away from a technology transfer approach that focuses on changing existing practice and toward a model that uses adjunct staff such as health educators—most often funded through outside grants—to do SBI for the organization. The use of adjunct staff to provide SBI is one way of adapting to the demands of funding agencies and practice guidelines that mandate the use of SBI. Though this method may be effective for SBI diffusion in emergency or trauma care settings, it is one that may not be transportable to other community settings wherein hiring additional staff such as health educators may not be feasible or appropriate. Within the mental health field, implementation of evidence-based practices is clearly the clinician’s charge (Drake et al., 2001).
Interest in mental health clinics as an alternative venue for SBI is growing, as evidenced by a county-wide SBI training for mental health providers in Los Angeles County in 2008. Mental health clinics offer several apparent advantages for SBI implementation. First, mental health clients are at increased risk for substance use disorders because of the co-occurring nature of substance abuse and mental health (Babor & Higgins-Biddle, 2000). Second, mental health clinicians already do counseling; it is their job. Primary care clinicians generally are not accustomed to counseling patients about substance use or using motivational interventions. Third, mental health clinicians are in a position to work with clients on an ongoing basis and can use the information from SBI in treatment planning.
UCLA Integrated Substance Abuse Programs received a 3-year SAMHSA grant in 2005 for the UCLA Access to Care project. UCLA was one of 12 campuses to receive a grant to implement a screening and brief intervention program to address binge drinking and drug use among college students.
Services were delivered at the UCLA Counseling and Psychological Services (CAPS), a multidisciplinary counseling center for the UCLA students. CAPS offers a variety of educational services such as outreach programs, wellness workshops, and wellness groups. It also provides clinical training programs, including APA-accredited pre- and post-doctoral internships. The current staff at CAPS includes 8 psychiatrists, 17 psychologists, and 6 clinical social workers. In the 2007–2008 academic year, CAPS served more than 6,000 students. Services at CAPS are paid for by university registration fees. Thus, CAPS does not face the challenge of reimbursement for SBI services, which is one of the major barriers to implementation of SBI in primary care and emergency care settings (McAvoy et al., 2001).
The ASSIST is an eight-question screening tool developed by the WHO in 1997 as a simple method of screening for hazardous, harmful, and dependent use of alcohol, tobacco, and other psychoactive substances (WHO ASSIST Group, 2002). The ASSIST covers the following substances: tobacco, alcohol, cannabis, cocaine, amphetamine-type stimulants, sedatives, hallucinogens, inhalants, opioids, and other drugs. Questions are asked about each drug the person has used, and substance-specific scores are derived at the end. The scores are interpreted as falling within a low-, moderate-, or high-risk range. The screen takes approximately 5 to 10 minutes to administer and provides information about the substances people have ever used in their lifetime, the substances they have used in the last 3 months, problems related to substance use, risk of current or future harm, dependence, and injecting drug use (for the ASSIST manual, see Henry-Edwards, Humeniuk, Ali, Poznyak, & Monteiro, 2003a).
The ASSIST has an accompanying brief intervention (BI) that is designed for individuals who score in the moderate- and high-risk ranges (Henry-Edwards et al., 2003b). The BI is based on the FRAMES model, which incorporates the following concepts: personalized feedback, responsibility, advice, menu of options, empathy, and self-efficacy. The BI uses a motivational interviewing approach and is designed to cover the following nine components: (1) giving feedback on one’s scores using the “report card” form; (2) giving advice about the health risks associated with use; (3) placing the responsibility for change on the individual; (4) expressing concern about the individual’s scores; (5) asking about the good things about using; (6) asking about the not-so-good things about using; (7) summarizing the conversation about the pros and cons of using; (8) expressing concern about the not-so-good things; and, finally, (9) offering take-home information. Educational handouts are given to patients at the end of the BI. The intervention takes on average 10 to 15 minutes to complete. Though individuals can score in the moderate-risk range for a number of different substances, clinicians typically focus on the substance with the highest score or any substance that a person reports injecting. The aim of the intervention is to help clients understand that their substance use is putting them at risk and to encourage them to reduce or give up their substance use. Other than the ASSIST screen, there are no other accompanying questionnaires that individuals need to complete for the brief intervention.
The process evaluation for the UCLA Access to Care project utilized multiple means to investigate the implementation of the ASSIST. Given the newness of the ASSIST to the practice world, our work was exploratory and focused on the “trialability” of the ASSIST. “Trialability,” or the ability to try out a new idea or practice before committing to it, is an important characteristic of innovations (Rogers, 1995). Given the exploratory nature of our project, we chose to use grounded theory (Glaser & Straus, 1967). We conducted qualitative field interviews with clinicians using the ASSIST and documented their comments in our ongoing meetings at CAPS.
As managers of the grant, our initial question was, “Are the clinicians conducting the ASSIST?” To track the use of the ASSIST, we extracted from clinic records the number of prescreens, positive results, and ASSISTs administered by clinicians. This activity was performed on a weekly basis. Owing to clinicians’ concerns about confidentiality, we were unable to audio-tape therapy sessions to discern fidelity to the ASSIST and BI. The following is a brief description of the research methods that we used.
To further document use of the ASSIST, project staff extracted information regarding procedures, SBI activities, and logistical barriers from our project meetings with clinicians. The research office was based at CAPS, and the project had a full-time research assistant who acted as a liaison between the clinic and the research staff. To document the experiences of clinicians, we conducted three focus groups and three individual interviews from April 2007 to July 2008. In addition, project meetings occurred on a regular basis with a core group of clinicians during the first year. Throughout the course of the 3-year grant, research staff met with the clinic director and clinical coordinator for the project on a monthly basis. Opportunities to share information with clinicians about the ASSIST occurred as part of the day-to-day activities at the clinic and in training sessions, lectures, and other project-related events.
Semi-structured interviews were conducted with 20 members of CAPS staff in three focus groups and three individual interviews. Fourteen of the participants were psychologists, four were licensed clinical social workers (LCSWs), and two were masters-level social work interns. There were eight female clinicians in the first focus group (six staff psychologists, one LCSW, and one psychology intern). The second focus group had four interns (psychology or social work), with one participant being male. The final focus group consisted of five psychologists and two LCSWs. There were six females and one male in this group. The individual interviews were conducted with a female LCSW, a male psychologist and administrator, and a female LCSW who was also the project clinical coordinator. Overall, there were more female than male clinicians on staff at CAPS (5 of 22 staff clinicians were male, not counting MDs).
The interviews took place at CAPS. All participants completed informed consent forms. The first author (Spear) conducted all interviews. Participants received a $20 gift card for their participation. The UCLA Office for the Protection of Research Subjects approved the study. The focus group discussions covered the following topics: how and when staff member administer the ASSIST; how they use the information in their work; strengths and weaknesses of the ASSIST; client reactions to the ASSIST and staff members’ response to using screening and brief interventions in a mental health setting.
All interviews were audio-taped and transcribed. Transcripts were first reviewed against the audiotapes to ensure accuracy and to obtain a general sense of common experiences and concerns. Codes were developed inductively as the transcripts were reviewed, allowing the data to dictate the analytical categories (Glaser & Straus, 1967). Common responses were highlighted and grouped, as were quotations that best illustrated the most common sentiments. Field notes from project meetings and observations were reviewed and used to triangulate our interview data. All interviews were conducted and analyzed by the first author (Spear); however, research and clinical staff were closely involved in interpreting the data and providing feedback. Research and clinical staff have collaborated in all phases of the project.
We designed our SBI procedures to target students who may be at risk and who could benefit from a brief intervention. Screening all students coming into the clinic with a 5- to 10-minute ASSIST was not feasible owing to the administration time. Instead, we used a four-question pre-screen with all students coming to the clinic to identify at-risk students. Students coming to CAPS for initial appointments or walk-in appointments and students who were mandated to come to the clinic because of an alcohol or drug-related violation completed the pre-screen before their appointments as part of routine intake paperwork. The pre-screen consisted of the three-question AUDIT-C (Dawson, Grant, Stinson, & Zhou, 2005) and one question that asked about any illicit dug use in the last 30 days (including any non-medical use of prescriptions). The AUDIT-C measures frequency and quantity of alcohol consumption and can be used to screen for active alcohol abuse and dependence. The question on illicit drug use read “Have you used any drug in the past 30 days not prescribed by a doctor?” “At-risk” was defined as a positive result on the pre-screen (the AUDIT-C or the illicit drug question). Receptionists at CAPS scored the pre-screen and flagged all positives in the students’ charts with a “post-it” note. For students who screened positive on the pre-screen, clinicians administered the ASSIST interview during the initial therapy session or in the follow-up visit. Clinicians, with the exception of psychiatrists, administered the ASSIST and provided the BI to students who scored in the moderate- and high-risk ranges. Students who scored in the high-risk range were offered referrals to a specialized treatment facility.
The initial plan was to have a core group of four clinicians administer the ASSIST and help with the overall procedures. The clinicians (three LCSWs, one psychologist) were selected by CAPS because of their interest in substance abuse and motivational interventions. These clinicians were thought to be the “early adopters.” In early 2006, we organized a day-long training on the ASSIST and BI for the core group and additional clinicians who were identified to eventually administer the ASSIST. The training was conducted by Bonnie McRee, MPH, and Janice Vendetti, MPH, of the Department of Community Medicine and Health Care, University of Connecticut School of Medicine. Clinicians learned to administer, score, and interpret the ASSIST and deliver the BI. Clinicians received an overview of motivational interviewing techniques. The training included didactic presentations and hands-on practice doing role plays. In addition to the initial training, new CAPS staff and interns were routinely trained by the clinical coordinator in the administration of these services as part of their orientation to CAPS.
In addition to the training, a licensed psychologist, who served as a consultant on the grant, provided on-site support a few hours each week at the clinic. Originally, the consultant was tasked with helping clinicians troubleshoot problems arising with the ASSIST and providing coaching on motivational interviewing and other components of the BI. However, the clinicians did not seek out the consultant’s assistance and, therefore, this approach was abandoned in the second year of the grant. It is possible that the therapists did not have time for an additional consultation or did not require assistance because they felt confident in their skills.
Not long after the implementation of the ASSIST screenings, it became clear that the small group of clinicians administering the ASSIST could not keep up with the demand. Over time, the core group of clinicians grew to about 10 clinicians. These clinicians regularly attended project meetings and were eager to administer the ASSIST. It became apparent, however, that many students whose AUDIT-C score met the threshold for receiving an ASSIST in their initial counseling session were not being screened with the ASSIST or receiving the BI. To increase the accessibility of SBI to students, a systematic change in clinic operations was required.
To help ensure that students received the ASSIST, the project made two pivotal decisions. First, the clinic director instituted a new policy that required all clinicians doing intake appointments to receive training on the ASSIST and begin using it. This policy increased the number of potential clinicians using the ASSIST from 10 to more than 20. Second, the clinic director decided to hire a clinical coordinator to help monitor the routine administration of these services and to devote several hours a day toward providing the screenings independently. This new clinician split her time between the Access to Care project and regular staff duties at CAPS.
Monitoring was conducted on a weekly basis by cross referencing the number of new clients who had scored positively on the AUDIT-C screening (computed by the front desk staff) and the number of completed ASSISTs for the week (extracted by the research assistant). When these numbers demonstrated a significant inconsistency, the clinical coordinator sent instant messages to clinicians who had not administered the ASSIST to a client reminding them to complete it at the next scheduled follow-up session. Additionally, the clinicians received a reminder on their clinical calendar to conduct the ASSIST at the client’s next scheduled session. Clinicians were also invited to refer any clients who flagged positively for an ASSIST directly to the clinical coordinator, who could then make arrangements to meet with the student independently to conduct the ASSIST and/or the BI at the student’s convenience.
Documenting the number of ASSISTs each week and reporting back to the staff was one of the activities of our research assistant. Overall, between February 2006 and October 2008, 6,786 students completed the pre-screen, and 38% of students screened positive (n = 2,569). Of the students who screened positive, 60% received the ASSIST screen and BI (n = 1,442). The discrepancy in number of positive pre-screens and of the ASSIST and BI administrations may be due to the fact that a large proportion of intakes were walk-ins, group intakes, and ADHD intakes. For these particular groups, ASSISTs were either less likely to be administered (for walk-ins owing to time constraints) or not required by definition-agreement (in the case of group intakes and ADHD intakes).
By having a dedicated clinician to provide the ASSIST and BI on demand, CAPS was able to conduct screenings either before or after a student’s initial appointment, to invite entire athletic teams to participate in the screening services, and to permit students to self-refer for the screening at their convenience. Also, dedicated time toward screening and BI services was routinely used to train new or temporary clinical staff (hired throughout each academic year) on how to provide these services and was also used to provide “booster” sessions for those who required additional consultation.
The original implementation plan called for administering the ASSIST during the initial intake appointment. In the field interviews, very few clinicians reported being able to accomplish this. Most clinicians reported administering the ASSIST in the second therapy session. The most common reason for not being able to administer the ASSIST in the first meeting was lack of time. Intake appointments are 50 minutes, and it was difficult for clinicians to find 15 to 20 minutes to conduct the SBI. Clinicians also tried to administer the ASSIST during emergency walk-in appointments. These appointments are a maximum of 20 minutes. All the clinicians we spoke with were unable to do the ASSIST in the context of a walk-in appointment.
There were other circumstances that prevented clinicians from doing the ASSIST in the initial appointment. One was couple’s therapy. Clinicians considered it awkward and inappropriate to do the ASSIST in the context of a couple’s initial therapy visit. The ASSIST was administered by some clinicians in couple’s therapy if one of the individuals screened positive. There was one case wherein the information that came out of the ASSIST helped a couple realize the role of drug use as a major factor contributing to their relationship problems. Clinicians also reported not conducting the ASSIST with students who were very distressed or who were suicidal.
When the lead researchers on the project (Spear and Rawson) originally proposed the SBI approach to CAPS, they assumed their use of the ASSIST would be a matter of adding additional questions on substance abuse to the existing mental health assessment. The idea was to substitute the standardized ASSIST questions for the one or two general questions that CAPS was already asking. In fact, the ASSIST, as a standardized interview, was very different for clinicians. In all three focus groups, the problem of adapting to the ASSIST was discussed. The main difference between the ASSIST and what they did before was the structured approach of the ASSIST. Clinicians were used to a free-flowing, unstructured approach to asking questions. In the past, there was no particular time during the initial intake appointment when substance abuse questions were asked, nor was this required or monitored by the clinic. With the ASSIST, the timing, format, and content of the questions were highly structured as compared to the interviewing approach clinicians were accustomed to.
In an effort to make the ASSIST blend with the rest of the clinical intake, clinicians reported that they tried to make the ASSIST more conversational by asking additional questions during the interview. The questions helped the clinicians get background information about the student. A few clinicians mentioned that they do not ask the questions exactly as they appear, to make them more conversational.
Despite clinicians’ unease with the structure of the ASSIST, they readily acknowledged that students were mostly accepting of the ASSIST and compliant to its use. Moreover, based on dialogue in the focus groups and project meetings, every clinician reported having some positive experiences with the ASSIST, such as when students expressed interest in the feedback and became more aware of their patterns of use. Comments such as “When it works, it works well,” “When it’s been relevant, I’ve been glad,” and “I’m glad we have it” were frequently heard in the focus groups and project meetings. Clinicians at CAPS are highly adaptive, as they admitted themselves, and discovered on their own that there were ways of introducing the ASSIST so that it felt (to the clinicians at least) like a natural part of the intake.
In project meetings, clinicians would demonstrate how they introduced the ASSIST by trying to “downplay” the formality of the interview. Clinicians who were the most comfortable with the process would say, “… Based on your earlier responses, I’d like to ask you some more questions about substance use.” Another strategy was to present the ASSIST as part of routine practice, which helped clinicians acknowledge the awkward transition from free-flowing dialogue to a standardized interview and “save face” because the ASSIST is something they have to do.
The implementation of the BI appeared to vary based on the amount of time clinicians had to give during the intake interviews, the students’ willingness to discuss their substance use, and the students’ presenting problem. For some clinicians, the BI appeared quite brief (e.g., 5 minutes) because of time constraints. The brief version consisted of giving the score, asking how students felt about their score, reviewing the health risks associated with substance use, and giving the educational handouts. When students did not see a problem with their use and were not willing to explore the topic, the motivational interviewing part of the intervention—the discussion of the pros and cons of using—became difficult. Most students did not report experiencing negative consequences from substance use on the ASSIST; however, clinicians reported making connections between substance use and mental health problems such as depression.
I just kind of relate it by saying … there’s just a big connection with mood and substance use, so I talk about that more as they could be someone who is anxious and they’re drinking or smoking pot. The drug use could be intensifying as a problem even if they are not identifying their use as a problem. So just pointing that out to them in a motivational interviewing way by saying, why don’t you track this and see what’s happening with your depression. And it just opens another way of talking about it. Sometimes you can see them glaze over and think, “Oh here we go with the alcohol and drug part.” But when you start linking it with mood and anxiety then they are like “Oh ok.”
Although some BIs were very brief, clinicians reported that sometimes the BI could be longer and even take an entire session if a student were presenting for a substance abuse–related problem, which was rare.
With respect to the “9 simple steps” framework for the BI, clinicians did not always follow the steps in order. Similar to their adaptations of the ASSIST, clinicians reported that they covered the required steps and content in their interventions but organized the steps in a way that fit with the overall flow of conversation. The one step that was the most awkward for clinicians was giving advice. To perform the “Advice” step, clinicians were trained to say, “The best way to reduce your risk is to cut down or stop using.” When clinicians were asked about how they implemented the BI, however, not one clinician spoke of giving advice. Giving advice about alcohol and drugs is not something that mental health clinicians usually do as part of therapy. Clinicians did report reviewing health risks with students and comparing students’ use with normative data (this information was presented on the educational handouts for alcohol, tobacco, and marijuana).
Clinicians highlighted a number of clinical benefits of the ASSIST. One of the most commonly reported benefits is that the ASSIST helps clinicians start a conversation with students about substance use. Before the Access to Care project, there were only a few questions about substance use on the clinic’s intake form, and most of the responses recorded on the intake forms were “problem denied” or “no problem reported.” The clinic director summed up the change in substance use assessment as a result of the ASSIST: “… Where we previously had clinicians simply avoiding probing this area, we now have clinicians who routinely offer the SBI as an important part of their broader treatment plan.”
It was not only that clinicians and students at CAPS were talking more about substance use than before, but the way of talking about substance use also changed. The ASSIST interview and BI encouraged a more direct way of talking with students. The use of structured talk required practice. By being required to do the ASSIST, most clinicians have gained significant experience with the intervention and report that, with time, they have learned to accept it and integrate it into their practice.
Clinicians with a more positive attitude about the ASSIST talked about the benefit of having a new skill to add to their repertoire. One of the interns explained,
… I had a vague idea on how to assess substance use but now I think I have a lot more knowledge in these other areas. I know what to look for and it is a way to give me a gauge to see if the person is at risk and how to approach them [about that risk].
The other benefit cited by clinicians was that the ASSIST works to build awareness among students about substance use and can help “shift their thinking” about their use. Clinicians build awareness by talking in a direct manner to students. In one example given by a clinician, the ASSIST provided a couple who were in denial about drug use the opportunity to discuss the issue of drug use. As the clinician recounted, “The ASSIST was the first direct talk about that.”
Clinicians reported that most students do not know about safe drinking limits. One of the most potent pieces of information that clinicians impart to the students is the connection between substance use and common social, academic, and psychological difficulties that students may face. Clinicians reported using the educational handouts, which often led to a more directed conversation.
One of the themes that emerged from the focus groups was that the ASSIST works well when it is relevant to the students’ concerns and presenting problems. One of the challenges that clinicians identified was having to do the ASSIST when a student reports little to no substance use. Clinicians commented in all three focus groups that the pre-screen resulted in too many false-positives. The following comment by one of the psychologists reflects a common sentiment among clinicians that the pre-screen sometimes identified students as being at risk when the clinicians felt otherwise. As she explained,
… a lot of people will say, yeah I smoked pot like once in the past 3 months and then you go ahead and do the whole ASSIST and that’s like the only thing. So it feels a little bit like you’re wasting your time and they’re looking at you like … what’s the big deal.
Several clinicians suggested that the threshold for a positive screen was too low. Clinicians talked a great deal about cases wherein students “flagged” as a positive, but substance use was not “an issue” or “not a treatment thing.” At the same time, some clinicians explained how the relevance of the ASSIST information had to be constructed or revealed by clinicians; the problematic nature of binge drinking or daily marijuana use was not obvious to many students in the context of an initial treatment session. The latter point was revealed on a number of occasions during the interviews, when the clinicians talked about the students’ underreporting their substance abuse or not being inclined to talk about substance use in the first session.
Underlying the question of the ASSIST’s relevance to treatment is a larger issue that has to do with clinicians’ therapeutic approaches. In the interviews, it became clear that there were two main orientations to understanding what should happen in intake appointments. These diverse orientations very much impacted the clinicians’ experiences with the ASSIST. For clinicians with a cognitive-behavioral approach, doing the ASSIST fit well. These clinicians were used to more structured interventions and comfortable with delivering educational information. They also appeared to have an easier time “switching gears” or moving from the topic of the presenting problem to substance use, which may or may appear to be related, at least not initially. One of CAPS administrators described it this way:
We always do educational and preventative mental health … You know someone’s coming in for something unrelated to sexual behavior but in the course of talking to them you hear a lot of risky stuff, you’ll move into that area even if they are here because they got busted stealing from the student union.
For psychotherapists who used a more humanistic or client-directed approach, the routine administration of the ASSIST screen and BI fit less well. Defining a problem for the client, controlling the flow of conversation, giving advice, discussing health risks, and using educational handouts are parts of the ASSIST that may not fit well with clinicians who come from a traditional psychotherapy background. For these clinicians, the ASSIST can take away from their goal of building rapport with their clients. One clinician recounted:
Assessment isn’t everything that’s happening in that first session. It’s sort of what is happening between the two of you relationally that’s gonna keep them coming back or have them follow-up on referral, so I don’t want to take the weight off of the assessment, it’s critical, but there’s this other piece, so any hesitation from clinicians is (because) you’re getting away from this other really important piece of joining and aligning.
Current SBI initiatives use a public health approach to reducing substance abuse and related health, social, and economic costs to communities. This public health approach tends to focus on population health and long-term benefits for populations and communities. As Babor and Higgins-Biddle (2000, p. 684) explain, “unlike clinical medicine, which is oriented almost entirely toward the patient’s presenting symptoms, the public health approach assumes that the impact of health promotion may go well beyond the individual patient.” It is within the context of this broader public health agenda that our project endeavored to engage mental health clinicians in the use of the ASSIST. It was a mission that researchers brought to the clinic. What the research staff did not quite appreciate until the project progressed was the extent to which implementation of the ASSIST would entail a shift in the therapeutic orientations of mental health clinicians as they became engaged in a community-wide prevention initiative.
Diffusion theory suggests that implementation of new practices happens in stages. Rogers (1995) outlined five main stages in his seminal work on diffusion of innovations: (1) knowledge, (2) persuasion, (3) decision, (4) implementation, and (5) confirmation. The first two steps involve the adoption of an innovation, which has to do with learning about an innovation and developing a favorable attitude toward its use. Steps 3 to 5 describe the implementation process, which entails making decisions about how to implement an innovation (e.g., strategic planning), putting it into practice and trying it out, and finally—if the innovation has proven useful—incorporating the practice into one’s behavioral repertoire. It is also assumed that individuals and organizations move through a similar process of change, one that focuses heavily on individual motivations and readiness to change.
In the Access to Care project, CAPS as an organization seemed to have moved through these stages in their implementation of the ASSIST. The organization adopted the idea of SBI and the ASSIST, in theory, during the planning phase of the project. In addition to the incentive of grant funding for the organization, the administrators saw the potential benefits of SBI for students and for the organization as a whole. CAPS administrators, after an initial trial period of a few months, instituted a policy requiring the use of the ASSIST among all clinicians (non-physicians). Implementation was built into the procedures of the clinic.
The adoption of the ASSIST and SBI by the administrators was reflected in how they spoke about the ASSIST. In the planning phase of the grant and when we first started training, the implementation of the ASSIST was talked about generally as something that was being done for the Access to Care project. Experience with the ASSIST was minimal in the initial year of the grant. Over time, the clinic director cultivated a clear message that framed the ASSIST as an important tool. Substance use was talked about more and more by clinicians as a critical problem warranting clinical attention. In a message directed to the clinicians, the director referred to the ASSIST as being “validated and effective in significantly reducing future harm associated with alcohol and drug abuse.” Whereas substance abuse may have been a peripheral focus before the grant, the director during the course of the project developed a message about substance abuse being a contributing factor to health and social problems, an “important piece of the puzzle.”
For the individual clinicians at CAPS, the diffusion process followed a different course than did the organization as a whole. Instead of moving through a gradual process of exposure, adoption, and implementation, clinicians entered the process at the implementation stage because of CAPS’ policy requiring the use of the ASSIST. Though a few key clinicians participated in the decision-making process to adopt the ASSIST, most clinicians were trained and simply started using the ASSIST. Clinicians’ attitudes toward the ASSIST were of less importance to ASSIST implementation than were the organization’s policy and ongoing monitoring by the clinical coordinator. It was clear from the focus groups that clinicians were doing this because it was required. Rather than making a decision about the tool up front after a training session or research presentation, it appears that acceptance of the ASSIST and the BI happened as a result of implementation, of working with the ASSIST over time. Several clinicians in the focus groups were honest about their dislike of the ASSIST initially. Over time, with practice and persuasion from management, clinicians began to incorporate the ASSIST into their repertoire, finding ways to adapt it to fit their personalities and therapeutic orientations.
Administering the ASSIST requires focusing at least 15 minutes of clinical time on one topic—substance abuse. As clinicians pointed out after the ASSIST became a required part of the intake process, no other topic gets treated like substance abuse at CAPS—that is, with a structured intervention and a feedback form. To make the ASSIST work, clinicians often “made it their own” by making the interview more conversational. One of the major adaptations the UCLA Access to Care project made to the ASSIST interview in 2008 was to develop an electronic version that students can complete on computers located in the clinic lobby. A pilot study of the computerized version has been completed (Spear et al., in review).
This study has some limitations. First, it was conducted with one clinic as part of a process evaluation for the UCLA ISAP Access to Care grant. By participating in a grant project, CAPS had a financial incentive to implement the ASSIST. Because the project actively promoted implementation, CAPS had exposure to researchers who provided research information and training resources. It is unclear whether such broad-based implementation by clinical staff would occur naturally in an organization that had no involvement with university collaborators. The work of the center is instructive in terms of the trialability of the ASSIST and the working implementation model that CAPS developed. It is worth noting that in 2008, CAPS staff trained 20 other university counseling centers in the use of the ASSIST. It would be worth exploring what motivating factors, organizational, and contextual circumstances provoked the interest in the ASSIST among these centers. The adoption and implementation processes of these centers would provide an opportune topic for a naturalistic diffusion study.
Second, interview data provide a snapshot of what people say about their experiences and activities but do not provide data about what people do in the actual context of implementation. Third, these data represent perceptions of one-half of the provider-client relationship—the providers’. Data on the experiences of students who completed the ASSIST would provide important evidence of its transportability on college campuses.
Future research could include a controlled trial of the ASSIST in mental health settings. Of particular importance is to better define the parameters of the BI. It is unclear at this point which components of the ASSIST BI are the most potent for reducing substance use risk behavior among students. The delivery of the BI can vary widely depending on the amount of time available during therapy sessions and students’ needs and interests. In addition, clinicians developed their own styles of administering the ASSIST and BI. It is important to determine how much flexibility clinicians have with the BI model so that adaptations can be made to fit the clinical environment.
An area for future research is the “reinvention” of the ASSIST and BI by therapists and how these changes affect client outcomes. Such a study would require a tool for monitoring the delivery of the ASSIST for fidelity to the current nine-step model. Last, a survey of student experiences with the ASSIST would be informative. Clinicians worried about the impact of a structured interview on their rapport with students and the possibility of losing students who may find in-depth questioning about substance use inappropriate or insensitive to their current needs.
Given the concerns raised by some clinicians in our field interviews, it seems appropriate to make some recommendations for future training on the ASSIST. In our conversations with clinicians, it appears that they could benefit from a discussion of the goals of SBI as a population-level prevention approach. Blending the ASSIST with therapeutic approaches can pose problems for some psychotherapists. The ways in which SBI can conflict with certain therapeutic approaches is another relevant topic for discussion among mental health therapists engaged in SBI work. Future training could also address the evidence to support brief interventions. In our regular trainings, little time was allowed for a review of the research evidence. In retrospect, that was deemed a mistake. A few clinicians questioned whether their short feedback sessions were sufficient for a BI and whether these sessions were effective. Although clinicians have limited time for training, spending some time on the research evidence is a worthy investment.
In conclusion, administration of the ASSIST in a campus mental health clinic is feasible. There are a number of benefits to using the ASSIST. Conducting the ASSIST screen provides an opportunity for clinicians to start a discussion with their clients about their substance use. In addition, the ASSIST can help clinicians discern whether substance use may be a contributing factor to the presenting mental health issue. As a result of the Access to Care project, many of the clinicians at CAPS reported that they learned new skills and developed a greater awareness of substance abuse problems. Adaptations such as a self-administered version may facilitate implementation and sustainability over time.
This study was funded by the Center for Substance Abuse Treatment of the Substance Abuse and Mental Health Services Administration (Grant TI 17244). The authors graciously thank the clinicians at UCLA Counseling and Psychological Services (CAPS) for their willingness to try the ASSIST and provide constructive feedback for the evaluation. The authors also wish to thank the administrative staff at CAPS for their ongoing support of the project and Kris Langabeer for her valuable editing assistance.