Treatments that clearly work are often not widely implemented in clinical practice. A plethora of published practice guidelines can frequently be more confusing than clarifying for the clinician. Questions on evidence-based practice, i.e. what kind of evidence, who decides what evidence should be implemented, and how do you do that, were addressed. Specific resources for accessing information about evidence-based practices are identified. The practical application of evidence-based practice guidelines developed by the Veterans Administration was offered. Dr. Hall-Flavin presented information on resources needed. Dr. Kivlahan reviewed the guideline development process. Dr. Hagedorn discussed four evidence-based practices in opioid agonist therapy as translated into practice.

Symposium Chairs: Mark Willenbring, MD, Minneapolis VA Medical Center, Minneapolis, MN and Daniel Hall-Flavin, MD, Mayo Clinic, Rochester, MN

What is evidence? Standard sources relied upon include empirical data, scholarly production, the library, the computer, and the laboratory. In psychiatry we think about other kinds of evidence, i.e., the truth that develops in interactions between people. Religious or spiritual experience in our field is regarded as a key component of practice. In listening to the debate about who decides what is best, scientists and clinicians have different ideas of what is best evidence. And, there is the question of what role the patient plays in determining what is best.

Practice guidelines are more likely to be valid if they are developed using systematic review. Involvement of regional or national groups from key disciplines further validates the guidelines. If explicit links exist between recommendations and scientific evidence, acceptance is supported in the field.

Randomized control trials (RCT), with their specific inclusion and exclusion criteria, result in 50-70% of the population being excluded from trials—these excluded individuals are our patients.

Clinical experience is based on how the physician was trained. We practice the way our colleague down the hall practices. Practice standards tend to be regional and local. We don’t seek out the latest RCT on the Internet. We are less likely to violate the norms of our local peers than we are to worry about the national norms. The last group of patients we saw influences us.

Increasingly, implementation of patient guidelines has to include patients. Patients are using the Internet to find information to tell physicians what they want.

Clinical experience is deceptive. In one study, 73% of primary care physicians believed that they immunized a greater percentage of patients than their colleagues did while they did not.

System barriers such as money, time, third-party payers rules, and problems with computers exist in implementation of best practice guidelines. Another problem is that we are all overwhelmed with information. Patients bring in information on the latest ad they read and ignore the physician’s advice. It is effective to give the patient the practice guideline to show what is necessary for optimal care.

Implementation strategies for a solid finding require broad skills, leadership and teamwork, negotiation and tact, selling ideas to peers, and linking evidence to the practice guidelines. What seems to be most effective is intensive conferencing with interaction and discussion and role-playing. Also pocket cards, wall charts, and pre-

formatted forms/reminders seem to have a moderate efficacy. We need to recruit opinion leaders to take an action to improve practice and then influence colleagues. We need to learn from the pharmaceutical industry about the use of academic detailing (outreach) to influence doctors.

The first step in implementing a practice guidelines is to model the process so the influences can be understood. The evaluation of efforts to implement a practice start with designing a logic model and creating a formative or implementation evaluation to assist in changing things right away. A process evaluation is needed to determine how much of the intervention actually occurred, and finally an outcome evaluation is required to determine if the desired event occurred.

In opiate agonist therapy, there are four specific practices that have a very good evidence base, i.e., (1) giving an adequate agonist dose, (2) adequate dose of counseling, (3) a maintenance versus detoxification or abstinence orientation, and (4) the use of contingency management. These clearly lead to decreased use of illicit opiates and increased program retention which in turn lead to remission of drug dependence, improved psychosocial and physical functioning, and decreased mortality.

The art of medicine means we deal with patients, not groups such as in RCTs. This is a problem when extrapolating from a RCT to an individual patient. So much of what we do in medicine is not based on RCTs.

To develop the practice guidelines, we needed an evidence summary. A panel of experts was created who identified a consensus statement and a dosing algorithm to assist clinicians in the evidence-based decision making process for an individual patient.

References:

1. Grimshaw J, Eccles M, et al. Developing clinically valid practice guidelines. J Eval Clin Pract. 1995;1(1):37-48.
2. Noe CA, Markson LJ. Prev Med. 1998;(6): 67-72.

Funded by Veterans Administration Quality Enhancement Research Initiative.

Evidence-based medicine (EBM) requires critical thinking. This presentation gave a perspective of evidence-based medicine as a practice modification that we can use to cover how to frame the question and access the literature. Why is evidence-based medicine important? New data may mean that new treatments are available, and there is a significant lag time from the discovery of something that may be beneficial to our patients and its incorporation into clinical practice.

Clinicians often find it difficult to find quality data that is new; our knowledge base naturally declines over time. Traditional CME does not necessarily improve clinical practice, and EBM emphasizes lifelong learning on how to practice and access quality information generated by others in a time-efficient manner.

EBM is defined as the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients. This implies the individual clinical expertise of the practicing clinician is based on accumulated experience and the efficient use of the best available external evidence.

Framing the question is the cornerstone to EBM. It is the formulation of pertinent answerable clinical questions that balance brevity with precision. The categories of inquiry include clinical findings, etiology, differential diagnosis, diagnostic testing, prognosis, treatment, prevention, and self-improvement—how do we learn from each individual patient encounter?

The well-built question has four basic principles. Starting with my patient, how would I describe a similar group? What main intervention am I considering with this patient? What is the main alternative (comparison intervention) to compare with this intervention? What could this exposure (clinical outcome) really affect? For example, in alcohol-dependent patients hospitalized for rehabilitation, does the addition of acamposate to naltrexone improve the outcome of treatment as measured by days abstinent compared to using naltrexone as monotherapy?

With so little time for practitioners, we need to prioritize in developing the questions. That is, which question is the most important for the patient’s well being; which is the most feasible to answer in the time available; which is most interesting to me; and which am I most likely to encounter repeatedly in my practice?

Medical literature expands in volume 7- 10% every year, doubling every seven to 10 years. There are approximately 20,000 medical journals and publications. Just in the field of internal medicine, there are 200+ monthly journals with over 500 editorials. Keeping up is almost impossible.

How do we cope with this volume of information? We check published reviews, go ask a colleague, attend a CME course, and consult practice guidelines, original research articles, and textbooks.

As use of CD-ROM databases increases, this may result in a decreased need for paper journals in the future. Quarterly updates are provided by CD-ROM databases. MEDLINE, an on-line database, is the most comprehensive one; www.PUBMED.com is a free service of the National Library of Medicine, which has comprehensive abstracts coming from the MEDLINE database as well as some evidence-based databases. Some institutions provide all employees with searchable databases on their terminals, i.e., Mayo Clinic provides MEDLINE, Current Concepts, American College of Physicians Journal Club, Best Evidence, and the Cochran database. Just typing in our search terms will search all these databases at once.

In setting your search strategies, first select a database specific to the field you are searching in and select studies that are scientifically strong and ready for use clinically. In selecting search terms use the National Library of Medicine’s MESH (Medical Subject Heading) categories. Use a combination of index and text words and be as specific as possible. Pare things down by using search terms such as AND, OR, and NOT. Spend some time with your medical librarian to learn the most efficient techniques for searches.

To cope successfully with the volume of literature, establish a surveillance strategy—choose a small number of core journals pertinent to your specialty for review as well as one or two general medical journals. Browse before reading in this order—first, articles pertinent to current practice, novel to the practice, with a relevant abstract; and second assess the author, study site and generalizability to your practice. Another source is the Journal for Evidence Based Mental Health. It abstracts the literature by taking high quality studies, digests them and provides summaries in an evidence-based format.

In September 2001, the Veterans Administration’s (VA) National Clinical Practice Guidelines Council endorsed the Veterans Administration/Department of Defense (VA/DoD) Clinical Practice Guideline for Management of Substance Use Disorders in the Primary and Specialty Care Settings. This presentation will provide an overview of the VA/DoD guideline cooperative development process. We consider these guidelines to make a substantial contribution to the field. If we knew then what we know now, the guidelines could have been produced faster, cheaper and with higher quality. In 1990, the Institute of Medicine defined clinical practice guidelines as "systematically developed statements to assist practitioner and patient decisions about appropriate health care or specific clinical circumstances." Guidelines are not a substitute for clinical judgement.

These guidelines have evolved since 1996. The collaboration with DoD was established in 2000, and the recommendation for approval by the VA National Guidelines Advisory Council received final signatures from the VA leadership this month.

The development process required recruitment of an expert working group, arrangement for technical support, clarification and re-clarification of goals and scope, selection of an algorithmic format, and specification and testing of the logic model for the approach. Other important steps in the process addressed the strength of evidence, translating the evidence into specific clinical recommendations, and explicitly grading the strength of those recommendations. Final steps prior to the approval included independent external review.

Guideline development is a team sport. The best players all have day jobs. My colleagues on the editorial panel, Drs. Willenbring, McNicholas and Miller, devoted enormous time and effort to the project. We also relied on a number of other AAAP experts during the process, as well as representatives from the DoD and other federal agencies. Multiple other disciplines were involved including nursing, psychology, social work, primary care medicine, and addiction counselors.

It was clear that the ultimate goal of the guidelines needed to be the improvement of patient outcomes through the improved quality of clinical services provided. Given the many adverse changes in VA specialty treatment since 1996, we were concerned about facilitating recognition and management of SUDs across the full continuum of federal health care, including primary and specialty health care settings.

The VA wanted to emphasize as a performance measure the mandate to improve screening in the primary care setting. The guidelines were organized into five modules:

1. Assessment and Management in Primary Care,
2. Stabilization (including Detoxification),
3. Assessment and Management in Specialty Care (Rehabilitation),
4. Addiction-Focused Pharmacotherapy, and
5. Care Management (which is not setting specific).

The VA had committed to an algorithmic approach to the guidelines. Selecting the format for the body of the guidelines took considerable trial and error.

The full guidelines contain the algorithms for each module along with annotation and evidence tables that support the guideline recommendations. Like other VA/DoD guidelines, we followed an algorithmic flowchart format. We recognized the caveat that the clinical reality is often not linear and requires attention to concurrent issues. This was tested as we followed the flow of simulated patients through the algorithm.

For each step in the algorithm there are annotations provided in the full guidelines. These identify the objective of that step, present the recommendations preferably using action verbs (ask, provide, assess, arrange), and give a brief rationale and type of supporting evidence. Recommendations for clinical decision making and actions were evidence-based to the extent possible.

The original literature review techniques for the guidelines would be much improved today with some of the techniques described by Dr. Hall-Flavin.

Evidence tables were developed that summarized recommendations, provided sources of evidence, included ratings of the quality of evidence, and graded the strength of the recommendation.

As you try to evaluate this guideline and other guidelines, I recommend that you review the AGREE (Appraisal Guidelines for Research and Evaluation) Collaboration. It is attempting to systematize the guideline review process in Europe. Their 23-item instrument covers six critical domains: scope and purpose, stakeholder involvement, rigor of development, clarity and presentation, applicability, and editorial independence. A growing concern as our process moves forward is to include the stakeholder, i.e., the patient.

Dissemination, implementation and revision of the guidelines are interrelated tasks. The electronic version is posted in downloadable files in DoD Web sites referenced below. Pocket guides were found to be an effective tool to get the guidelines into daily use. Guideline revisions are critical. Paul Shekelle recently reviewed 17 guidelines and concluded that the average time in which they became obsolete was 5.8 years. He recommended validity be reviewed every three years. I submit that in addiction treatment our pace is even faster than that. Revisions should be based on treatment benefits/harms, outcomes considered and important, available interventions. The revisions should rely on:

1. indications that current practice is optimal,
2. the values placed on outcomes, and
3. available resources.

References:

1. Cluzeau FA, Littlejohns P, Grimshaw JM, et al. Development and application of a generic methodology to assess the quality of clinical guidelines. Int J Qual Health Care. 1999;(1):21-8.
2. Shekelle P, et al. When should clinical guidelines be updated? BMJ. 2001;323(7305):155-7.
3. Shekelle PG, Ortiz E, Rhodes S, et al. Validity of the Agency for Healthcare Research and Quality clinical practice guidelines: how quickly do guidelines become outdated? JAMA. 2001;26;286(12):1461-7.
4. Appraisal of Guidelines Research and Evaluation Web site. Available at: http://www.agreecollaboration.org. Accessed March 5, 2002.
5. Center for Health Care Evaluation Web site. Veterans Affairs Health Care System, Palo Alto, CA. Quality Enhancement Research Initiative. Available at: http://www.chce.research.med.va.gov.queri. Accessed March 5, 2002.
6. National Guideline Clearinghouse Web site. Sponsored by the Agency for Healthcare Research and Quality. Available at: http://www.guideline.gov. Accessed March 5, 2002.
7. Clinical Practice Guidelines. Office of Quality and Performance, Veterans Health Administration. Department of Veterans Affairs. Available at: http://www.oqp.med.va.gov/cpg/cpg.htm. Accessed March 5, 2002.
8. Quality Management Office. U.S. Army MEDCOM. Available at: http://www.cs.amedd.army.mil/qmo/Home.htm. Accessed March 5, 2002.

This presentation describes a Facilitated Quality Improvement Demonstration Project. We recruited eight VA agonist clinics into the project. Our team is available to assist them with best practices for opiate agonist therapy.

Available evidence strongly suggests the efficacy of specific opiate agonist therapy practices. The practices include adequate dosing and counseling, maintenance orientation (keep in treatment for a long period of time versus goal to stabilize and detoxify), and contingency management. These target practices lead to decreased use of illicit opiates and increased program retention. Patients stay in the programs longer. This leads to increased rate of remission of drug dependence, improved psychosocial and physical functioning, and decreased mortality.

For an adequate opiate agonist dose targeted practice, higher dose means less opiate use. With a low dose of (40-50mg), the positive rate was a little over 60% and in the high dose patients receiving 80-100mg, the positive rate was in the low 50%. Higher dose also means fewer dropouts from the program.

For the adequate counseling frequency practice, I reviewed McLellan’s study in which he compared three levels of psychosocial services, i.e., no counseling (methadone only), standard counseling (weekly x four weeks, then minimum of biweekly), and enhanced social services (standard counseling plus additional social services). The enhanced patients had better outcomes than the patients in the standard counseling group, and the patients in the standard group had better outcomes than the no-counseling group on all outcomes. Sixty-nine percent of the no-counseling group had to be protectively transferred into the standard counseling group because their outcomes were so poor. This was defined by having three crises that required medical attention or they had constant positive urines without any indication in change in use.

The standard counseling was the most cost effective. It paid for itself in the reduction of use of emergency services and other crisis interventions for the patients. While extra money went into the enhanced counseling, it was good for the patients but was not cost effective.

Maintenance orientation has become one of the more controversial practices as far as the clinics are concerned. Caplehorn compared program retention for a strongly abstinence-oriented (Clinic 1) program to a more maintenance-oriented (Clinic 2) program across four six-month periods. Patients in Clinic 1 were twice as likely to leave treatment against staff advice in the sixth to 12^th months of treatment; and were three times as likely to leave treatment in the 12-18^th months of treatment. Clinic 1 was a two-year treatment program.

Contingency management practice is a novel concept. Stitzer, who is consulting on this project, did the study I reviewed. Patients were either assigned to a contingency management protocol in which the patient received a take home privilege—take the dose of methadone home on one day rather than have to come every single day to dose. This was granted after two consecutive weeks of drug-free urine. In the noncontingency protocol, the patients received their take-home dose randomly. The results after six months showed that 32% of the patients that received the contingency management protocol were classified as improved. They were able to put together four consecutive weeks of abstinence. When the random group was switched into the contingency management group, 28% of the switchers were able to put together four weeks of clean time.

We advise our clinics to provide take home privileges on the basis of two weeks worth of clean urine screens. This is revoked immediately if the patient comes up with a positive after that. The behavioral goal is simple. With this objective, the patient knows exactly what they have to do, there is no question, and there is no negotiating with the staff. The reward is more immediately tied to the behavior goal. Someone could potentially get a take home dose two weeks into the program.

Under current practices in the VA for dosing, about 40% of opiate agonist patients are on stable doses of <60mg. There is a significant portion of those patients that are stable and on adequate doses. About 30% of opiate agonist programs in the VA routinely recommend detoxification to every patient. The counseling availability varies considerably across programs. So far, everyone has at least monthly counseling, but it ranges from weekly for every single patient (no matter how long they have been in the program) to monthly sessions for every patient.

Our transfer team goals are to work with participating programs to improve implementation of these four areas to decrease illicit opiate use and increase program retention. These goals were chosen for clinic ease in monitoring outcomes. Both outcomes based on the research literature should be directly affected by changes in program practices. Research evidence indicates these outcomes are strongly predictive of later outcomes—decreased mortality and increased psychosocial well being of the patient.

Our study uses the VA method of Plan, Do, Study, Act (PDSA). The question is first asked whether it is worth looking at the current practice. This approach supports a continuous look at whether the original change has been effective and incorporates plans for a new and continuing PDSA cycle.

References:

1. Capelhorn JRM, et al. Clinic Policy and Retention in Methadone Maintenance. International J Addict. 1993;28(1):73-89.
2. Hamilton and Humphreys. Program Evaluation and Resource Center; 1996.
3. McLellan et al. The effects of psychosocial services in substance abuse treatment. JAMA. 1993;269(15):1953-9.
4. OpiATE Initiative. Evidence Summary. Use of Contingency Management in Opioid Agonist Therapy. VA Affairs Medical Center, Minneapolis, MN.
5. OpiATE Initiative. Evidence Summary. Counseling Frequency in Opioid Agonist Therapy. VA Affairs Medical Center, Minneapolis, MN.
6. OpiATE Initiative. Evidence Summary. Dose of Opioid Agonist. VA Affairs Medical Center, Minneapolis, MN.
7. OpiATE Initiative. Evidence Summary. Abstinence vs. Maintenance Orientation in Opioid Agonist Therapy. VA Affairs Medical Center, Minneapolis, MN.
8. Strain EC, et al. Moderate vs. high-dose methadone in the treatment of opioid dependence: a randomized trial. JAMA. 1999;28(11),1000-5.
9. Stitzer ML, et al. Contingent take-home incentive: Effects on drug use of methadone maintenance patients. J Counsel Clinical Psych. 1992;60(6):927-34.