Treating substance abuse in primary care: a demonstration project

Purpose The purpose of this project was to implement the delivery of a full range of substance abuse services in a primary care setting. Implementation and logistical issues including confidentiality and communication are discussed. The delivery of services, types of patients, and contextual and policy factors that influenced project implementation are described. Context Substance use disorders are associated with significant morbidity and mortality worldwide. Patients with alcohol/drug problems frequently present in primary care. Effective and brief treatments are available and recommended for primary care but infrequently implemented. Institutional and provider barriers to implementation have been identified. Data source Project documentation, data from the patient tracking system, and clinical case notes were used for description and analyses. Conclusion Addressing substance abuse problems in primary care is important. Behavioral health professionals with training in substance abuse can provide a range of services that are likely to enhance the quality and quantity of care available to patients. Although contextual factors needed to be addressed, integration of services was manageable and seemed acceptable to both providers and patients in this project.


Introduction
Substance abuse profoundly affects public health worldwide. According to the World Health Organization w1x, 4% of the burden of disease and 3.2% of all deaths globally are attributable to alcohol. Only about half of all alcohol-related problems are caused by people with alcohol dependence w2x. The Institute of Medicine w3x recommends that the medical system address the full spectrum of alcohol use disorders. Primary-care-based treatment for other drug problems is at an earlier stage, but effective brief interventions have been documented w4x and buprenorphine has now been approved for office-based treatment of opioid dependence w5x.
There are good reasons to attend to substance abuse in primary care. Alcohol problems are overrepresented in many populations seeking medical care, affecting up to 44% w6x of primary care patients. Tobacco smoking is often addressed in primary care, and brief interventions for other substance abuse are effective w7-15x, available w16-19x, and recommended w3, 20, 21x though seldom integrated into routine care. Although at-risk drinkers are often reluctant to seek specialist addiction treatment, about two-thirds do visit their general practitioner each year w7x. The prescription of dependence-producing medications also warrants attention to potential abuse. Addressing substance abuse within the context of primary care could increase access to and retention in care w22-24x, and improve health outcomes w25, 26x. Providers have preferred integrated care for managing alcohol problems w27x, and potential societal benefits include reduced costs of health care and social problems related to substance abuse, particularly for patients with chronic medical and mental health disorders w28-30x.
Why, then, is substance abuse care so seldom integrated into routine health services? Common barriers include system and reimbursement issues, lack of provider training and knowledge, and challenges in identifying and engaging patients w17, 31, 32x. Providing training and ongoing support for providers can somewhat enhance care w33, 34x, though institutional factors such as competing priorities and time demands continue to inhibit successful implementation by physicians w35x.
An alternative for integration is to use physicianextenders to deliver behavioral health services within primary care. In clinical trials, medically-trained providers including nurses, physician assistants, and residents, successfully delivered a brief outpatient psychosocial treatment for alcohol dependence, using naltrexone, an emphasis on abstinence, and medical case management w24, 36x. Another option, referred to as a Primary Mental Health Model w29x utilizes behavioral health specialists to deliver brief assessment and treatment in health care settings under the direction of the primary care provider. A comparison of a specialist-delivered versus primary care providerdelivered brief interventions found that specialists could be more effective and less costly, depending on provider and institutional characteristics w27, 37x.
Generalist and specialist services within primary care are not mutually exclusive. Substance abuse treatment services vary along a continuum of intensity. Brief advice, medication management, case management, and certain brief treatments can be effectively delivered by generalist practitioners within a primary care structure. Indeed, it has been argued w38x that a strength of primary care is in the use of generalist skills to address the broad range of health care issues related to substance abuse. More intensive outpatient services include detoxification treatment provided by addiction specialists. Bringing these together, it is possible for specialists to deliver substance abuse services within primary care settings. The United Kingdom Models of Care for Alcohol Misuse w39x recommends including specialist services as a second tier of interventions.
Integrating care also includes strong linkages between specialty treatment and primary care. Addiction treatment has occurred primarily in specialist settings, without ongoing primary care. Such integration can include expedited referral arrangements shared practice guidelines, and a coordinated continuum of levels of care w40x. The specific nature of these linkages will differ by level of urbanization w41x or organization of care delivery w42x, population served w43x, type of financing available w44x and national or regional policies w45x. Past linkages have focused mostly on bringing primary care to patients in addiction treatment w22, 26, 30, 46x and facilitating referral to specialist treatment by primary care providers w23, 42x.

Description of the care procedure
The purpose of this demonstration project was to implement delivery of a full range of substance abuse services within a primary care setting. The goal was to explore the feasibility and acceptability of these services to providers and patients in a busy urban primary care clinic serving a culturally diverse population. Objectives included offering support and expertise to staff and providers, providing on-site treatment for patients, building linkages with specialist treatment facilities, and documenting the process. We specifically explored system barriers and the types of primary care patients needing substance abuse services.
The project was a collaborative effort of the University of New Mexico's Center on Alcoholism, Substance Abuse and Addiction with its Department of Community and Family Medicine, and was funded by a substance abuse innovation grant from the Robert Wood Johnson Foundation. The site was the Family Practice Clinic, a teaching clinic providing primary care for patients regardless of income. Though social workers and case managers were on staff, the clinic had neither routine screening nor systematic interventions for substance use disorders. Ordinary care consisted at most of asking whether patients drank and how much, treating alcohol-related medical problems, and occasionally suggesting specialist treatment.
For 20 months, the first author, with master's level training in clinical psychology and substance abuse, was placed as an on-site behavioral health counselor at the clinic 20 hours per week. She was charged with developing 'from scratch' a system for providing evidence-based practices for the screening, evaluation and treatment of substance abuse or addiction problems identified during routine primary care. The first several months required resolving logistical issues, educating and negotiating with staff and providers about the program. She gradually became a member of the treatment team, had access to medical records, and served as a resource for the attending physicians, primary care providers, and pharmacy, social work and case managers. She shared an office with the case managers and saw patients in regular clinic rooms.

Confidentiality and communication
Substance abuse services in the United States are subject to the Code of Federal Regulations Title 42 w47x, which protects and regulates the documentation and communication of patient information related to treatment for substance use disorders. These regulations do not address integrated services such as those provided in this project. Legal counsel determined that the program was subject to the regulations because the services were specific to substance abuse, and psychotherapy was provided to some patients. This required keeping separate therapy notes that were not part of the open medical record. Brief summary reports of visits were kept in a secure section of the medical record with extremely limited access, as the primary method to document patient care and facilitate communication between providers. The clinic was just converting to an electronic medical record that had not yet been modified to safeguard such sensitive, regulated information. Therefore, care summaries were kept on paper, not scanned into the electronic medical record and thus less accessible to the primary care provider. Fortunately, the first author's regular presence in the clinic allowed for onsite communication of pertinent information related to the treatment provided.

Scheduling and space
Scheduling of clinic time and space were critical concerns for this project. A centralized scheduling system for primary care providers allocated specific clinic rooms during their scheduled hours. Support providers were not in the centralized scheduling system, controlled their own schedules, and negotiated for space, which was frequently hard to find. For the first year of this project, the behavioral health counselor was considered a support provider. This meant that patients could not schedule appointments except by direct contact. This obstacle was eventually eliminated by adding the counselor to the centralized scheduling system.

Consultations
The behavioral health counselor interacted with clinic providers in several ways. Providers could request immediate consultation prior to a patient primary care visit, invite the counselor to collaborate in delivering a brief intervention during the visit, andyor walk the patient to her office for further consultation after the medical visit. She also, in response to formal referrals, scheduled specialist on-site visits with patients, sometimes in tandem with follow-up medical care. Most often the referring provider introduced the program to the patient, did not personally participate in the consultation, received a report from the counselor, and collaborated to coordinate follow-up care.
The content of substance abuse consultations centered on the delivery of evidence-based treatments w48x. The overall clinical style was motivational interviewing w49, 50x, a patient-centered and goal-directed approach to facilitate behavior change by resolving ambivalence about change. The counselor also drew on a menu of cognitive-behavioral strategies, relapse prevention, behavioral contracting, and case management as needed, similar to the Combined Behavioral Intervention developed for and tested in the multisite COMBINE trial w51x.
Referrals were facilitated by a significant clinic policy change that was implemented during the project, requiring providers to obtain random urine drug screens for all patients prescribed narcotics or other highly addictive medications. Patients testing positive for illegal drugs, including marijuana, received a warning that a second positive test would result in terminating their prescription. Patients with positive drug screens were referred for consultation, sometimes as they were being tapered from medications. These included chronic pain patients with long histories of high-dose narcotics. Others were referred because they had developed substantial tolerance to their pain medications. Treating these patients required a broader range of behavioral health care. Most of these patients would not consider referral to a specialist addiction treatment program, but did engage in onsite consultations. Some were using illicit drugs for pain control, particularly when they experienced severe side effects from or tolerance to narcotic medications. In such cases, the counselor helped patients to manage depression, anxiety, and pain. This illustrates the advantage of having a more broadly trained behavioral health specialist with expertise beyond substance abuse.

Patients served
During the project period, a total of 89 different patients (69% female) were scheduled, with a mean age of 40 years (range 15-77). Observed ethnicity was 69% non-Hispanic Caucasian, 24% (21) Hispanic and 6% (5) African-American. Most patients (79%) were referred by their primary care provider, with the rest referred by a nurse, social worker, clinical pharmacist, family member or self-referred. Most were treated for problems associated with alcohol (38%), illicit drugs (34%), both alcohol and illicit drugs (10%) or prescription drugs (9%). The remainder was treated This article is published in a peer reviewed section of the International Journal of Integrated Care for gambling (2), smoking (3) or seeking help for a family member's substance abuse (3).
A total of 516 visits were scheduled. Of these, 111 (22%) were missed, with 14 patients (16%) never attending. Most visits (76%) were conducted in person, with the first typically done at the time of the primary care visit, and 24% were completed by telephone. About one third of treated patients had either one visit or -1 hour of treatment. This is comparable to brief intervention recommended for primary care settings. Another third of the patients received 2-5 visits or 1-4 hours total. This is consistent with effective brief treatments such as motivational interviewing. The remaining patients received more intense treatment in both number of visits and total time. Their treatment consisted of case management, follow-up telephone calls, check-in during primary care visits, and facilitation and management of referrals. Six patients were referred out for additional treatment.
Of the 23 patients who received only brief intervention, all but one had been scheduled for only one session, usually by the patient's choice. The other scheduled but failed to come for further appointments. These were opportunistic interventions, in that they had not been seeking help for the problem discussed: alcohol, drugs, smoking or gambling.
Six patients had significant legal, social and family problems related to their substance abuse and were required by authorities to seek treatment in order to retain benefits, stay out of jail, or recover custody of their children. These patients also had medical issues, including chronic illness, disability or follow-up care for hospitalization or injury. In these cases the counselor provided short-term 'bridge treatment' that met the system requirements until referral to specialist treatment was completed. This required coordination with the legal and social systems as well as community programs.
A small but important group was the two adolescents seen. They had intact families who were concerned about their use of alcohol and marijuana, and had a long-term relationship with their primary care provider. The adolescents were high functioning and their involvement with drugs and alcohol was in early stages. Providing counseling in the primary care setting reduced stigma for the families, normalized the process of seeking help before problems become severe, and invited the adolescents to rethink their substance use in a safe environment.
Finally, 33 patients had substance use problems of varying severity related to medical care issues. Some referrals were triggered by significant risk factors including pregnancy, elevated blood pressure or liver enzymes, or family and social problems reported by the patient or family members. For some patients, substance use had precipitated a medical or psychiatric crisis that had been resolved but required followup care. These crises included overdose, severe detoxification, sexually transmitted diseases, severe abscesses as a result of skin popping, hospitalizations for infections, or other serious medical conditions such as dehydration and loss of blood. These patients were medically stable and sought substance abuse consultation to prevent recurrence. Several of these patients requested or were offered medications such as disulfiram or naltrexone to support their efforts to manage their drinking.
Of these 33, 20 patients had chronic conditions including hepatitis C, asthma, epilepsy, pancreatitis, chronic pain and diabetes, all of which require consistent monitoring, frequent visits, and a team of providers for good disease management. As such medical conditions worsen, patients' lives increasingly center on the illness, and substance abuse treatment becomes a small yet significant part of the overall picture. Having substance abuse treatment integrated into routine chronic disease care reduced the burden and kept substance use on the agenda. The term dual-diagnosis (referring to having both a substance use disorder and a mental health diagnosis) does not begin to capture the extent of comorbidity and the severe decrease in functioning of this group. Most patients carried one or more major medical diagnosis, a significant psychiatric diagnosis (including depression, anxiety, bipolar disorder or schizophrenia), and a diagnosed or undiagnosed substance use disorder. In addition, many had documented personality disorders, developmental delays or severe disabilities. Many had a long history of substance abuse and mental health treatment, involvement with social service and legal systems, and poverty.

Discussion
It just makes sense to address substance use in primary care. Most people with substance use disorders are already being seen by primary care providers for other concerns, many of which are related to or exacerbated by alcoholydrug problems. Effective treatments are available, many of which can be delivered within the context of primary health care.
There are at least three models for addressing substance use in primary care. One is for providers to offer relatively brief interventions in the course of ordinary care. Materials to use with patients and to guide brief counseling are readily available w16-19x and there is strong clinical trial evidence that brief This article is published in a peer reviewed section of the International Journal of Integrated Care intervention can reduce alcohol, tobacco, and other drug use w52x. The availability of effective medications also enhances the options for managing substance use disorders in primary care w36x. For alcohol problems disulfiram, naltrexone and acamprosate were all available at the time of the project, but seldom prescribed. A drug for treating heroin addiction, bupenorphrine, that can be prescribed and managed in primary care settings, also became available during the project. The main constraints in physician management of substance use disorders seem to be lack of training and the time pressures of medical practice.
A second option is referral of patients to addiction treatment clinics. This may be particularly appropriate for patients with severe dependence. Significant obstacles here may be lack of available and affordable services, and patient reluctance to seek formal addiction treatment. Many specialist treatment programs, however, are not well prepared (or funded) to treat patients with less severe alcoholydrug problems, which would constitute the majority of those seen in primary care. During this project, access to specialist substance abuse treatment was severely limited for poor, uninsured patients. Even the most accessible public programs had long waiting lists. The major public psychiatric facility was overwhelmed, essentially accepting only patients in extreme crisis. Patients with dual diagnosis were often passed back and forth depending on which diagnosis seemed primary. Some patients had been involved in both mental health and substance abuse treatment and perceived that they had been 'kicked out' and were not welcome back. It thus fell back to the primary care providers to manage their medical problems as well as their psychiatric and substance use disorders.
A third option already being implemented in Europe and highlighted in this article is to have behavioral health specialists offer services on-site within primary care systems. Most evidence-based treatments for substance use disorders can be offered as outpatient consultation in a primary care clinic. Providing such services in the context of healthcare can reduce stigma and increase patient access to appropriate treatment. This approach also provides for closer coordination of specialist consultation with primary care, as would be desirable in the treatment of any chronic illness. An ideal provider would be a behavioral health professional who is competent in treating substance use disorders and who can also provide consultation on other psychosocial and behavioral problems that arise in relation to healthcare. Such professionals can be valuable team members in caring for multi-problem patients.
There are some obvious obstacles to co-locating behavioral health and primary care. One is that most providers and patients are still unaccustomed to such one-stop service in primary care. Another is that it can be challenging to find behavioral health professionals with sufficiently broad training. U.S. addiction counselors have historically focused solely on alcoholydrug problems, and may be unaccustomed to working with patients having less severe use, problems and dependence. Many mental health professionals, on the other hand, have received little or no training in how to address substance use disorders w53x. Behavioral health professionals may also be unaccustomed to communicating and working collaboratively with primary health care providers. There are practical issues to be resolved around credentialing, scheduling, reimbursement, space, documentation, confidentiality, and interdisciplinary communication and collaboration.
The use of any of these three models suggests that there should be routine screening for alcoholydrug problems in primary care, much as providers routinely ask about smoking. Even a single question such as ''How often do you have four or more drinks in one day?'' can pick up a substantial proportion of people with hazardous drinking. Similarly, one can ask how often patients use illicit drugs, with follow-up questions to any answer other than ''never.'' Add these to a question about tobacco use, and you have in three questions a screen that is far better than no screening at all. Asked in a matter-of-fact manner, such questions can also detect use and problems at much earlier stages of development, while prevention and treatment are easier. Without routine screening, one is likely to detect only more severe and entrenched substance use problems.

Conclusion
It is a surprisingly new idea that substance use disorders fall within the domain of primary health care, at least within the United States where alcoholydrug problems have historically been thought of as a separate issue to be dealt with only by specialist services. Consequently, substance dependence is the only chronic illness for which there has been no primary care. There are persuasive reasons to integrate the treatment of these disorders within mainstream healthcare. There are various models for doing so, and effective treatments are available. It is overdue that routine screening and services for substance use disorders should find their rightful place within primary healthcare.