Addressing Substance Abuse in Health Care Settings William R. Miller, Catherine Baca, Wilson M. Compton, Denise Ernst, Jennifer K. Manuel, Beverly Pringle, Carol R. Schermer, Roger D. Weiss, Mark L. Willenbring, and Allen Zweben This article summarizes the proceedings of a roundtable discussion at the 2005 annual meeting of the Research Society on Alcoholism in Santa Barbara, California. The chair was William R. Miller. The presentations were as follows: (1) Screening and Brief Intervention for Alcohol Problems, by Allen Zweben; (2) Three Intervention Models and Their Impact on Medical Records, by Denise Ernst; (3) Pharmacotherapies for Managing Alcohol Dependence in Health Care Settings, by Roger D. Weiss; (4) The Trauma Center as an Opportunity, by Carol R. Schermer; (5) Motivational Inter- viewing by Telephone and Telemedicine, by Catherine Baca; (6) Health Care as a Context for Treat- ing Drug Abuse and Dependence, by Wilson M. Compton; and (7) Interventions for Heavy Drinking in Health Care settings: Barriers and Strategies, by Mark L. Willenbring. Key Words: Brief Intervention, Motivational Interviewing, Screening, Healthcare A LTHOUGH ONLY A small minority of people with alcohol abuse or dependence seek specialist treatment, most are seen each year in health care or social services. Furthermore, alcohol use disorders are overrepresented among patients in most health care and social service pop- ulations and are associated with disproportionately poor outcomes. Yet, alcohol problems are seldom screened for or addressed systematically in these settings. There are good reasons to address alcohol problems within routine health and social services settings (Miller and Weisner, 2002). Patients often decline referral to spe- cialist treatment, but will accept consultation from current providers. Effective cross-culturally validated tools are available to screen for alcohol use disorders in general service settings. There is strong evidence for the efficacy of relatively brief interventions (BIs) for alcohol problems, which often yield outcomes similar to those for more extended specialist treatment (Babor and Grant, 1992; Chapman and Huygens, 1988; Moyer et al., 2002; Project MATCH Research Group, 1997). Effective pharmaco- therapies further enhance the feasibility of managing substance use disorders in general practice settings (Swift, 2003). All of this highlights the importance of addressing sub- stance abuse and dependence in the course of regular health care. This series of presentations summarizes the recent findings, models, and challenges in doing so. SCREENING AND BRIEF INTERVENTION FOR ALCOHOL PROBLEMS Allen Zweben Individuals with alcohol problems are often encoun- tered in high-volume health care settings. Within this con- text, there is an opportunity to address alcohol problems before they become worse. Screening and BI can be effec- tive in managing alcohol disorders within a health care setting, particularly with individuals having low-moderate alcohol problems. This presentation is focused on the 3 following areas: (1) useful screening tools for identifying at-risk individuals, (2) effective components of BI, and (3) strategies to surmount obstacles to providing screening and BI in health care settings. Useful Screening Tools There are several considerations in choosing suitable screening tools for health care settings. First, individuals seen in health care settings are often unaware of having an alcohol problem, particularly those with few consequences associated with their drinking practices. These individuals are being seen for medical concerns and consequently may be confused or may react negatively if asked about their drinking practices, even though alcohol may play a salient role in their current health problems. These negative reac- tions may be related to the stigma customarily attached to having a ‘‘drinking problem.’’ Consequently, it would be an advantage to nest alcohol screening within a more general health habits interview. This allows one a greater oppor- tunity to discuss alcohol use in the language of health promotion and thereby reduce the stigma associated with the ‘‘alcoholism’’ label. It also helps patients make the link- age between their medical condition and their alcohol use. Other considerations involved in choosing screening methods are as follows: (1) sensitivity and specificity of From the University of Mexico Center on Alcoholism, Substance Abuse and Addictions (CASAA) (WM, CB, DE, JM); the National Institute on Drug Abuse (WC, BP); the University of North Carolina School of Medicine (CS); Harvard Medical School (RW); the National Institute on Alcohol Abuse and Alcoholosm (MW); and Columbia University School of Social Work (AZ). Received for publication September 30, 2005; accepted October 17, 2005. Copyright r 2006 by the Research Society on Alcoholism. DOI: 10.1111/j.1530-0277.2006.00027.x Alcohol Clin Exp Res, Vol 30, No 2, 2006: pp 292–302. 292 ALCOHOLISM:CLINICAL AND EXPERIMENTAL RESEARCH Vol. 30, No. 2 February 2006 PDF Create! 5 Trial www.nuance.com
Transcript
Addressing Substance Abuse in Health Care Settings
William R. Miller, Catherine Baca, Wilson M. Compton, Denise Ernst, Jennifer K. Manuel,Beverly Pringle, Carol R. Schermer, Roger D. Weiss, Mark L. Willenbring, and Allen Zweben
This article summarizes the proceedings of a roundtable discussion at the 2005 annual meeting ofthe Research Society on Alcoholism in Santa Barbara, California. The chair was William R. Miller.The presentations were as follows: (1) Screening and Brief Intervention for Alcohol Problems, byAllen Zweben; (2) Three Intervention Models and Their Impact on Medical Records, by DeniseErnst; (3) Pharmacotherapies for Managing Alcohol Dependence in Health Care Settings, by RogerD. Weiss; (4) The Trauma Center as an Opportunity, by Carol R. Schermer; (5) Motivational Inter-viewing by Telephone and Telemedicine, by Catherine Baca; (6) Health Care as a Context for Treat-ing Drug Abuse and Dependence, byWilson M. Compton; and (7) Interventions for Heavy Drinkingin Health Care settings: Barriers and Strategies, by Mark L. Willenbring.
ALTHOUGH ONLY A small minority of people withalcohol abuse or dependence seek specialist treatment,
most are seen each year in health care or social services.Furthermore, alcohol use disorders are overrepresentedamong patients in most health care and social service pop-ulations and are associated with disproportionately pooroutcomes. Yet, alcohol problems are seldom screened for oraddressed systematically in these settings.There are good reasons to address alcohol problems
within routine health and social services settings (Millerand Weisner, 2002). Patients often decline referral to spe-cialist treatment, but will accept consultation from currentproviders. Effective cross-culturally validated tools areavailable to screen for alcohol use disorders in generalservice settings. There is strong evidence for the efficacy ofrelatively brief interventions (BIs) for alcohol problems,which often yield outcomes similar to those for moreextended specialist treatment (Babor and Grant, 1992;Chapman and Huygens, 1988; Moyer et al., 2002; ProjectMATCH Research Group, 1997). Effective pharmaco-therapies further enhance the feasibility of managingsubstance use disorders in general practice settings(Swift, 2003).All of this highlights the importance of addressing sub-
stance abuse and dependence in the course of regularhealth care. This series of presentations summarizes therecent findings, models, and challenges in doing so.
SCREENING AND BRIEF INTERVENTION FOR
ALCOHOL PROBLEMS
Allen Zweben
Individuals with alcohol problems are often encoun-tered in high-volume health care settings. Within this con-text, there is an opportunity to address alcohol problemsbefore they become worse. Screening and BI can be effec-tive in managing alcohol disorders within a health caresetting, particularly with individuals having low-moderatealcohol problems. This presentation is focused on the3 following areas: (1) useful screening tools for identifyingat-risk individuals, (2) effective components of BI, and (3)strategies to surmount obstacles to providing screeningand BI in health care settings.
Useful Screening Tools
There are several considerations in choosing suitablescreening tools for health care settings. First, individualsseen in health care settings are often unaware of having analcohol problem, particularly those with few consequencesassociated with their drinking practices. These individualsare being seen for medical concerns and consequently maybe confused or may react negatively if asked about theirdrinking practices, even though alcohol may play a salientrole in their current health problems. These negative reac-tions may be related to the stigma customarily attached tohaving a ‘‘drinking problem.’’ Consequently, it would bean advantage to nest alcohol screening within amore generalhealth habits interview. This allows one a greater oppor-tunity to discuss alcohol use in the language of healthpromotion and thereby reduce the stigma associated withthe ‘‘alcoholism’’ label. It also helps patients make the link-age between their medical condition and their alcohol use.Other considerations involved in choosing screening
methods are as follows: (1) sensitivity and specificity of
From the University of Mexico Center on Alcoholism, SubstanceAbuse and Addictions (CASAA) (WM, CB, DE, JM); the NationalInstitute on Drug Abuse (WC, BP); the University of North CarolinaSchool of Medicine (CS); Harvard Medical School (RW); the NationalInstitute on Alcohol Abuse and Alcoholosm (MW); and ColumbiaUniversity School of Social Work (AZ).
Received for publication September 30, 2005; acceptedOctober 17, 2005.Copyright r 2006 by the Research Society on Alcoholism.
ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH Vol. 30, No. 2February 2006
PDF Create!
5 Trial
www.nuanc
e.com
the particular measures, (2) kinds of populations receivingthe service, and (3) the time and resources available toconduct the screening interview. For example, in settingswith very limited resources, a single question on drinkingpractices may suffice. One such question (‘‘Whenwas the lasttime you had more than __ drinks’’) was able to detect 86%of individuals with an alcohol disorder (Fleming, 2003).Table 1 gives examples of commonly used alcohol screeningmeasures, the populations for whom the instruments weredesigned, and their advantages and disadvantages.
Effective Components of Brief Intervention
Brief intervention has been well established as an effec-tive approach for reducing alcohol use with individualswho are considered to be nondependent drinkers. Theessential components of BI include (1) screening, (2) feed-back on personal risk, (3) advice-giving on changingdrinking behavior, (4) self-help information for changingdrinking practices, and (5) referral for formal help, if nec-essary. The overall goal of BI is to enable patients to makethe connection between various health and social problemsand their alcohol use patterns, thereby facilitating a com-mitment to change.More recently, motivational interviewing strategies have
been incorporated into the components of BI (Hettemaet al., 2005). Motivational interviewing complements theother components, particularly with clients who are ambi-valent about recognizing and accepting the need to changetheir hazardous drinking patterns (Miller and Rollnick,2002). Brief interventions have been used opportunisticallyas a secondary prevention approach, in primary care clin-ics and emergency rooms of hospitals.
Despite advances made in refining and testing BI, muchis still unknown about the mechanisms of change associat-ed with the approach. Also, further research will be neces-sary to determine the long-term effects of BIs and theeffectiveness of the approach with certain ethnic or agegroups (e.g., elderly or adolescent population).
Surmounting Obstacles to Providing Screening and BriefIntervention in Health Care Settings
Failure in health care settings to provide screening andBI for alcohol problems can be attributed to a number ofobstacles including (1) practitioners’ attitudes towardtreating alcoholic patients, (2) absence of resources avail-able to deal with alcohol problems, and (3) lack of stafftraining in detecting, intervening, and referring individualswith alcohol problems.Being unfamiliar with screening and BI technology such
as standardized screening tools and motivational interview-ing strategies, together with patients’ defensiveness (i.e., lowreadiness to change), can lead to a sense of helplessness onthe part of practitioners in dealing with alcoholic patients.Evidence has shown that general practitioners who wereprovided with skill training in screening and BI techniqueswere more likely to use such techniques in routine clinicalpractice than those without such training.In addition, institutional factors such as the lack of staff
time, inadequate financial incentives, or having a generallyunsupportive environment for intervening with alcoholicpatients may contribute to low practitioner motivation toconduct screening and BI. Amulticomponent strategy thatinvolves increased institutional support combined withpractitioner training could help change the delivery of
Table 1. Alcohol Screening Measures
MAST (Michigan alcoholismscreening test)
Adults andadolescents
Psychometric properties tested in numerous studies Time referent problems; less useful withindividuals not seeking help for alcoholproblems
CAGE (Cutting down, annoyingyou, guilt, and eye-opener)
Adults andadolescents
Readily incorporated into a health screening interview Less sensitive with African American men,white women, and Mexican Americans
More sensitive than CAGE or MAST with AfricanAmerican women
Warrants further testing
CHARM (cut down, how to use,annoyed, reasons for useand more than intended)
Older adults Easily incorporated into a motivational-style briefintervention; deals with both alcohol problems andprescription drugs
Warrants further testing
Biological Markers - GTP(g transferace) and CDTcarbohydrate deficienttransferrin)
Adults Not affected by cognitive or motivational factors;useful as ‘‘adjunctive aid’’ (i.e., combined withstandardized screening tools) for identifyingharmful consumption
Sensitivity and specificity affected by age,gender, smoking status, medicalproblems, and medication use
293HEALTH CARE SETTINGS
PDF Create!
5 Trial
www.nuanc
e.com
alcohol treatment services in health care settings. Providingfinancial incentives and improving linkages or integra-tion with specialty settings for referral adherence would bea beginning step in the right direction. Also, hiring of spe-cialists to work closely with busy physicians in implement-ing screening and BI protocols would go a long way inimproving alcohol treatment services in health care settings.
THREE INTERVENTION MODELS AND THEIR IMPACT ON
MEDICAL RECORDS
Denise Ernst
Intervention for substance abuse problems in health caresettings can be delivered in a variety of ways. Consider thefollowing 3 models: (1) minimal assessment and referral tospecialty services, (2) brief intervention provided by thehealth care provider in the context of medical care, and (3)specialist expertise colocated in the health care setting. Thefirst of these has been the dominant model in most medicalsettings. Specialist in-depth assessment and treatment havegenerally been provided by separate addiction treatmentfacilities. A major limitation to this model is that for var-ious reasons, few referred patients actually enter specialisttreatment. The second model, using health care providersthemselves to intervene and provide advice orother brief interventions, has demonstrated efficacy in thetreatment of problems associated with substance use (Milleret al., 2003). However, the model is not widely disseminatedin practice. Reported barriers include lack of time, reim-bursement, and training (Fleming andManwell, 1999). Ear-ly studies integrating addiction treatment with primarymedical care have demonstrated better outcomes in termsof the addiction and overall health (Friedmann et al., 2003;Willenbring and Olson, 1999a, 1999b). The third model—integration of behavioral and mental health services intoprimary care—is considered critical to establishing a coor-dinated, comprehensive stepped-care approach to treatingmental health, including addiction (Blount, 1998).The stigma associated with substance use disorders and
receiving treatment for these problems prompted Congressto enact legislation in the early 1970s to protect thepatient’s right to confidentiality in treatment, known as42 CFR (Code of Federal Regulations Title 42—PublicHealth: Chapter I—Public Health Service, Department ofHealth and Human Services, Subchapter A—General Pro-visions; Part 2—Confidentiality of Alcohol and DrugAbuse Patient Records, Subpart A—Introduction, Cur-rent through June 17, 2002). This legislation has guidedpractice in US substance abuse treatment programs sincethat time. It prevents the sharing of any information out-side of the treatment program without the explicit permis-sion of the patient except under very specific situationssuch as the need for emergency medical care. It also pro-tects that information from being shared with the sponsorsof the health care such as employers, courts, or insurancecompanies. Even within the substance abuse program,
information is shared on a limited need-to-know basis.These protections were put in place to make treatment a‘‘safe’’ option for patients. In the mid-1980s, the legislationwas amended to exclude emergency hospital services. Othermedical providers and systems may still be subject to theregulations if they have a substance abuse program in place.The 3 models listed above are affected differently by
42 CFR. In the first model, simple referral to specialisttreatment for those patients whose substance abuse prob-lem is discovered in the course of regular care is not cov-ered under these regulations; there are no special require-ments for the medical records in these practices. However,the substance abuse treatment program to which referral ismade most likely would fall under 42 CFR, and thus with-out explicit permission from the patient, the providerreceives little to no information concerning the outcomeof the referral. Primary care providers (PCP) often per-ceive this process as sending patients off into a black hole,and coordination of care is limited.In the second model, whereby health care providers
deliver BIs in the context of primary care, there are againno special patient protections (under 42 CFR) beyondstandard HIPAA regulations. Information about patientsubstance abuse, which may be highly stigmatizing, be-comes part of the regular medical record, accessible to allwho have access to the chart.When a specialist substance abuse provider works with-
in a general health care setting, the 42 CFR regulations doapply, and the entire program or practice becomes a pro-vider of those services. In this case, the confidentiality ofpatient records regarding substance abuse must be pro-tected, communicating only essential information to theprimary care or referring provider. The specialist isallowed to keep separate therapy notes that do not becomepart of the medical record, and access to these notes isclosely restricted. The problem here is determining what isessential for the primary care provider to know. Primarycare providers themselves may argue that they need toknow everything in order to provide the best care to thepatient. Any substance abuse notes placed in the medicalrecord are likewise protected, with access limited to thosewho need to know (another definitional complexity). Ingeneral, this information is not released without the spe-cific permission of the patient or with a court order. Mostpractices must modify their medical record system toaccommodate this extra level of protection. It also requireseducation of providers, assistance with documentation,and efforts to reduce stigma for the patient.There are hybrid models in which the primary care envi-
ronment does not fall under 42 CFR. If a behavioral healthspecialist working in a primary care setting provides briefservices for a variety of psychological and behavioral issues(which may include substance abuse) working under thedirection and treatment plan of the PCP, this is not con-sidered specialist treatment and separate records are notkept (Strosahl, 1998). Such programs have usually designed
294 MILLER ET AL.
PDF Create!
5 Trial
www.nuanc
e.com
medical record documentation systems to standardize howthe information is recorded and to minimize patient stigma-tization. Such mainstreaming of substance abuse within inthe overall context of health care can itself be destigmatiz-ing. This hybrid model is also comparable with a stepped-care approach for addressing hazardous drinking that maynot be appropriate for specialist treatment.Substance use disorders are highly prevalent and can
have devastating effects on patient health and on society.Medical systems have a unique opportunity (and somewould say, a moral obligation) to address these issues. Inchoosing a model, programs and providers should ensurethat patient confidentiality is protected whether or not theprogram falls under 42 CFR. Health care–based interven-tions can screen for and address not only problems relatedto alcohol, but illicit and prescription drug abuse as well.With such major effects on population health, substanceabuse belongs within the domain of concern for PCP.
PHARMACOTHERAPIES FOR MANAGING ALCOHOL
DEPENDENCE IN HEALTH CARE SETTINGS
Roger D. Weiss
Although most studies of pharmacotherapy for alcoholdependence have been carried out in specialized alcohol-ism treatment programs, general health care settings offera unique opportunity to treat patients’ alcohol problems aspart of their ongoing health care. The advantage of pro-viding pharmacotherapy for alcoholism in primary caresettings is the ability to treat a much larger number ofpatients than would enter specialty treatment, includingindividuals with a much wider continuum of severity ofalcohol problems. Specialty treatment centers typically at-tract individuals with relatively severe alcohol problems,whereas people with even relatively mild alcohol problemsare seen in primary health care settings.One of the characteristics of alcoholism pharmacother-
apy in health care settings is the use of relatively brief,medically oriented counseling that accompanies the pre-scription of medication. This counseling approach, whichhas been described in published manuals (Pettinati et al.,2004; Volpicelli et al., 2001), typically includes ongoingmonitoring of drinking: support and encouragement, dis-cussion of medication side effects, an emphasis on theimportance of medication adherence, monitoring of over-all health, and recommendations for support group (e.g.,Alcoholics Anonymous) and/or counseling attendance.Not all medically oriented counseling occurs in primarycare health care settings, however. For example, some re-search has shown promising results for alcoholism phar-macotherapy in conjunction with brief, medically orientedcounseling in specialty settings (Johnson et al., 2003).Two types of studies have examined alcoholism phar-
macotherapy in health care settings. The first type ofresearch examined whether pharmacotherapy for alcoholismcan be effective when accompanied by medical counseling,
without specialty psychosocial treatment. The secondtype of study investigated whether medication to treatalcoholism can be successfully incorporated into primarycare settings.O’Malley et al. (2003) conducted the first type of study,
which examined the use of naltrexone in a primary caresetting. In this study, 197 participants all received open-label naltrexone and 10 weeks of either primary caremanagement (delivered by a nurse practitioner, a physi-cian assistant, or a physician) or cognitive-behavioral ther-apy (conducted by a psychologist or social worker). Thoseindividuals who responded successfully to either primarycare management or cognitive-behavioral therapy werefollowed for an additional 24 weeks, receiving eithernaltrexone or placebo in a double-blind fashion.The investigators found no significant difference in pos-
itive drinking outcomes (i.e., the avoidance of persistentheavy drinking) between the primary care managementpatients and those receiving cognitive-behavioral therapy,although the latter group was better able to maintainabstinence. Interestingly, however, those who respondedto the primary care management model were more likelyto maintain their beneficial response with naltrexone thanwith placebo; this was not true for the cognitive-behavioraltherapy responders. This study thus showed that healthcare practitioners can achieve good results with the use ofnaltrexone and a medically oriented counseling approachin a primary care setting.The second type of research, examining the addition of
alcoholism pharmacotherapy to a primary care setting, isillustrated by a study with acamprosate (Kiritze-Toporet al., 2004), which recruited 149 French general practi-tioners, who treated 422 alcohol-dependent patients witheither ‘‘standard care’’ or standard care plus open-labelacamprosate (2 g/d) for 12 months. The primary outcomemeasure for this study was the change score for the alco-hol-related problems questionnaire (ARPQ; Patienceet al., 1997). At the end of the study, those patients whoreceived acamprosate showed, on average, one less alco-hol-related problem on the ARPQ, a greater duration ofcumulative abstinence, and only a 17% dropout rate.Thus, this study showed the potential benefit of addingalcoholism pharmacotherapy to general medical care inthe treatment of patients with alcohol problems.
Research-to-Practice Cautions
Although the 2 studies described above demonstrate thepotential benefit of using pharmacotherapy for alcoholismin primary care settings, it is important to note that certaincharacteristics of these studies may limit their generaliz-ability. For example, the O’Malley et al. (2003) studyrecruited participants from advertisements and from analcoholism treatment center. Kiritze-Topor et al. (2004)cautioned that their general practitioners, because of theirhigh level of interest in alcoholism, may not be represent-
295HEALTH CARE SETTINGS
PDF Create!
5 Trial
www.nuanc
e.com
ative of most physicians. For pharmacotherapy of alco-holism to be successful in general health care settings, pri-mary care physicians need to screen patients for alcoholproblems and intervene when appropriate. Unfortunately,this process has not always been successful (Beich et al.,2002), but the historical example of other disorders givesrise to optimism; both depression and erectile dysfunctionbecame part of mainstream medical treatment wheneffective and easily prescribed pharmacotherapies becameavailable. With the recent introduction of a thirdFDA-approved medication for alcohol dependence, andongoing studies of other medications, it is possible thatpharmacotherapy for alcoholism will become part of themainstream of general medical practice.Acknowledgments: From the Department of Psychiatry,
Harvard Medical School, Cambridge, Massachusetts; andthe Alcohol andDrug Abuse Treatment Program,McLeanHospital, Belmont, Massachusetts. This work was sup-ported by Grant U10 AA11756 from the National Inst-itute on Alcohol Abuse and Alcoholism and Grant K02DA00326 from the National Institute on Drug Abuse.
THE TRAUMA CENTER AS AN OPPORTUNITY
Carol R. Schermer
Alcohol has been recognized as a major risk factor forinjuries for more than 4 decades, and injuries from motorvehicle crashes are the leading cause of alcohol-attributa-ble deaths (MMWR, 2004). The increased risk of healthproblems and injury associated with excessive alcohol con-sumption results in an overrepresentation of heavy drink-ers in medical settings. Because hazardous and harmfuldrinkers comprise a larger segment of the population thandependent drinkers, most of society’s alcohol problems areattributable to nondependent drinkers.More than 30 randomized trials demonstrate the effica-
cy of brief counseling interventions for hazardous andharmful drinkers (Moyer et al., 2002). For these nonde-pendent, but still hazardous and harmful, drinkers, the useof screening and BIs may reduce the risk of future healthproblems and injury and may also reduce the risk of thedevelopment of more severe and difficult-to-treat cases ofalcohol problems (Bien et al., 1993; Gentilello et al., 1999;Wilk et al., 1997). Increasing evidence shows that BIs foralcohol disorders effectively reduce alcohol consumption,recurrent injury (Gentilello et al., 1999; Longabaugh et al.,2001), motor vehicle crashes (Mello et al., 2005), and driv-ing under the influence (Longabaugh et al., 2001; Montiet al., 1999; C. R. Schermer et al., unpublished data, 2005).Based on epidemiologic, social science, and medicalresearch, we have concluded that trauma centers representan excellent opportunity for alcohol screening and briefinterventions (ASBIs) because of the following:
1. Nearly one-half of injuries are alcohol related (Rivaraet al., 1993).
2. Trauma patients who are problem drinkers are at highrisk for recurrent injury if their alcohol problemremains untreated (Gentilello et al., 1999; Schermer etal., 2001).
3. Trauma patients have often never been screened fortheir alcohol use (Schermer et al., 2003a).
4. People admitted to trauma centers are highly likely tobe hazardous and harmful drinkers (Rivara et al., 1993;Soderstrom et al., 1997).
5. Trauma patients are willing to be screened for and totalk about their alcohol use (Schermer et al., 2003a).
6. Trauma surgeons support the idea of screening and BIs,and the body that oversees trauma center designationnowmandates screening and intervention for level 1 andlevel 2 trauma centers (Schermer et al., 2003b).
7. Even at very busy trauma centers, a half-time employeeis sufficient to perform screening interventions (Scher-mer, 2005).
As part of a project funded by the Robert Wood John-son Foundation substance abuse policy research program,we surveyed trauma surgeons about alcohol screening andinterventions and also evaluated the feasibility of perform-ing ASBIs in trauma centers by implementing an empiri-cally based intervention into 3 trauma centers naı̈ve toalcohol screening and intervention. The purpose of ourtrauma surgeon survey was to determine their support forimplementing ASBIs in trauma centers. We asked aboutcurrent alcohol screening and treatment practices in theirown trauma centers, their particular knowledge andbeliefs, and what they believed were barriers to screeningand intervention. The results of the survey (Schermer et al.,2003b) showed that the overwhelming majority of sur-geons (83%) agreed that a trauma center is an appropriatesetting to address a pattern of harmful alcohol consump-tion. In addition, most surgeons (86%) agreed that it wasimportant to talk to injured patients about harmful alco-hol consumption. However, only 25% of surgeons usedformal screening questionnaires to identify patients withalcohol use disorders. Nearly one-half of surgeons (49%)reported that they understood the concept of BI, but only5% of surgeons were comfortable negotiating behaviorchange with their patients. Moreover, 59% reported thatfewer than half of the patients with suspected alcoholproblems discuss their alcohol use while at the trauma cen-ter. The survey items on ‘‘perceived barriers to screeningand intervention’’ showed that only 2% of surgeonsthought screening and counseling would significantlyincrease health care costs; 7.6% thought screening wastoo time consuming; and 14% thought it might compro-mise patient confidentiality. However, 27% perceivedscreening as a threat to reimbursement.The main outcome measure of the survey was whether
the surgeons endorsed the ideas of screening and BI pro-grams. They were asked to respond to the following 2statements on the survey: (1) I support the idea of alcohol
296 MILLER ET AL.
PDF Create!
5 Trial
www.nuanc
e.com
screening with injured patients, and (2) I support the ideaof BI for injured patients who have alcohol problems.Overall, 88% of surgeons supported screening and 72%supported BIs. A multivariate analysis revealed that sur-geons who supported screening were also 6.2 times morelikely to believe that a trauma center is an appropriate set-ting to address alcohol disorders than those who did notsupport screening. Surgeons were 5.7 times more likely tosupport BI if they understood its concepts. Compared withsurgeons who did not support BI, those who supported itwere about 7 times less likely to think that screening andintervention would increase costs too much. These data ledus to believe that we needed to inform surgeons about theconcepts of screening and BI and to show them what thecosts of screening and BI would be.Another part of the project was to implement an alcohol
screening and intervention program in 3 trauma centers todetermine time constraints, whether hospital employeestrained in screening and brief intervention could deliverthe intervention, and whether hospitalized patients wouldaccept the program. Although these centers employedtrauma surgeons with extensive clinical and research expe-rience, there had been no prior research or clinical interestin alcohol screening or interventions. To implement theprograms, 1 half-time research assistant position wasfunded at each of the 3 centers. The research assistantshad no prior experience performing alcohol screening,motivational interviewing, or brief interventions, but theyreceived 1.5 days of training in alcohol screening andmotivational interviewing to become an ‘‘interviewer.’’They were trained by a surgeon and a psychologist, bothof whom are members of the motivational interviewingnetwork of trainers (www.motivationalinterview.org).Specifically, the interviewers were trained to use the alco-hol use disorders identification test (AUDIT) as thescreening instrument and were trained in an adaptation ofmotivational interviewing for the intervention.Trainers taught an interview style similar to that de-
scribed by Rollnick et al. (1999) in the bookHealth Behav-ior Change: A Guide for Practitioners. They emphasizedthe following 3 strategies for engaging patients in a non-judgmental discussion of drinking:
1. Discussing pros and cons. An easy, nonthreateningway to start a discussion is to ask patients whatthey like and dislike about their drinking experiences.
2. Examining importance and confidence rulers. A methodof asking patients to rank 2 questions on a scale of 1 to10.How important is it to change your drinking habits? Ifyou decided to cut down or stop drinking, how confidentare you that you could do it?
3. Empathy. The use of reflective listening as amethod of por-traying an understanding of what the patient is saying.
Even with just a half-time interviewer, we found thattime constraints did not impede the screening and inter-
vening for eligible patients. On a typical day the averagepercentage of patients ineligible for screening because ofseverity of illness (i.e., mechanical ventilation or closed-head injury) ranged from 30% to 75%. On weekdays,despite very busy clinical trauma services, the half-timeinterviewers, who worked 5 d/wk, were able to screen allpatients and offer interventions to appropriate patients.However the interviewers at all sites were unavailable onweekends. This resulted in between 15 and 20% of patientsnot being screened. Moreover, language barriers variedsubstantially among the 3 sites, rendering a number of pa-tients ineligible for screening in our project but arguing fora need for multilingual interventionists.Importantly, the vast majority of patients at sites B and
C readily accepted screening and intervention; only 3% to8% of patients declined screening when it was incorporat-ed into the daily routine. Although no outcomes weretracked and no patient identifiers were collected, 1 of the 3sites mandated informed consent and a signature to par-ticipate in the project. At that site, the screening declinerate was nearly 20%.Once screening was complete, patientswith an AUDIT score of 8 or greater were offered a30-minute BI. The interviewer posed a simple questionsuch as, ‘‘Would now be a good time to talk with you alittle bit more about your drinking?’’ Or the transition wasperformed using a statement such as, ‘‘Tell me a little bitmore about your drinking.’’ Of more than 100 patientsoffered BI, only 1 declined talking with the interviewer,and that was because the patient stated he preferred todiscuss his alcohol use with a formal counselor.Program implementation was tested under busy circum-
stances, as 2 of the 3 trauma centers admit more than 4,000patients/y. The study demonstrated that BIs can be imple-mented using interviewers who had no prior knowledge ofmotivational interviewing or BIs and as few as 1.5 days oftraining. One half-time employee, working daily, was ableto screen most patients.
Conclusions
Trauma centers represent an excellent opportunity foralcohol interventions. Because alcohol misuse is so preva-lent in trauma centers, and because untreated alcohol dis-orders place individuals and society at great risk for futureinjuries and hospitalizations, trauma centers should screenand intervene for alcohol disorders. Most trauma surgeonsalready support ASBIs. Our findings showed that anytrauma center committed to setting up a program can suc-cessfully do so with a relatively small investment. Toencourage widespread implementation, it may be neces-sary to further educate the trauma community that screen-ing and BI programs are effective, that standard methodsalready exist, the cost of implementation is low, and pa-tient acceptance is high.Acknowledgments: Research and manuscript prepara-
tion were supported in part by the Robert Wood Johnson
297HEALTH CARE SETTINGS
PDF Create!
5 Trial
www.nuanc
e.com
Foundation Substance Abuse Policy Research ProgramGrant 044119 and by the National Institute on AlcoholAbuse and Alcoholism K23 AA-00310.
MOTIVATIONAL INTERVIEWING BY TELEPHONE AND
TELEMEDICINE
Catherine T. Baca and Jennifer K. Manuel
When long distances from counseling services are aproblem, could long-distance communications be a solu-tion? For people with alcohol problems, long distancesfrom substance abuse counseling services can be a sub-stantial barrier to appropriate treatment.Alcohol misuse is widespread among rural adults and
adolescents in the United States (Robertson et al., 1997);yet limited treatment options are available in rural areas.Rural areas account for a large proportion of allalcohol-related motor vehicle accident injuries and fatali-ties (Zwerling et al., 2005). At the same time, 20% of ruralcounties have no mental health services (Gamm et al.,2003). Specialty mental health providers are still fewer percapita in rural areas than in urban areas (Goldsmith et al.,1997; Merwin et al., 1995). This disparity is even worse inrural minority communities (Mueller et al., 1999). Tele-health (interactive audio–video) and telephone (interactiveaudio) sessions are potential solutions to the dearth of treat-ment available in rural areas. Research is lacking in the useof telehealth for substance abuse counseling. One studyfound the telephone useful for continuing care for bothalcohol and cocaine dependence (McKay et al., 2005).We have recently conducted a pilot study comparing
client satisfaction with 3 different modes of communicatingin motivational interviewing (MI) (Miller and Rollnick,2002). Thirty participants responded to an Albuquerquenewspaper advertisement for people concerned about theirown drinking. Inclusion criteria for the study were as fol-lows: (1) 18 years of age or older, (2) sixth grade readinglevel, (3) positive AUDIT screen (threshold), and (4) writ-ten consent. exclusion criteria were as follows: (1) severeillness (e.g., acute psychosis) that would interfere withtheir ability to understand consent, assessment, and inter-vention procedures; (2) inability to participate in the fol-low-up assessment; and (3) alcohol counseling in theprevious 3 months.We randomized participants to one of 3 groups as fol-
lows: (1) telehealth, (2) telephone, or (3) face-to-face. Thetechnology used for telehealth was H.320, which enabledhigh-quality interactive (real-time audio and video) using aswitched ISDN circuit. Two MI sessions were offered toparticipants in each group. One session was offered imme-diately after the pretreatment assessment, and the secondsession was offered no later than 2 months after the first.Two counselors trained in MI delivered therapy across the3 groups. The MI followed the clinical style describedby Miller and Rollnick (2002). Of central importancefor counselors in this study were empathy, frequent use of
verbal reflections, open-ended questions, elicitation ofchange talk, attention to verbal discrepancies between cli-ent values and behavior, enhancing self-efficacy, and roll-ing with resistance in a nonconfrontational manner.Follow-up satisfaction assessments were collected after
each MI session. The participants were asked about whichof the 3 modes of MI delivery they would have preferred,their satisfactionwith eachMI session, and their satisfactionwith the counselor. Given a choice of only the 2 long-distance communication modes (telephone or televideo),participants were then asked which their preferred modewould be. Participants reported high satisfaction with all 3modes ofMI communication and, in a pairwise choice, pref-erred telehealth (video) contact to audio (telephone) alone.At least in terms of patient satisfaction, it appears that
long-distance delivery modes of MI can be successfullyimplemented for alcohol counseling of problem drinkers.Overall, communication with the 3 modes of MI commu-nication was well received by participants, with some pref-erence to be able to see the counselor, even on video.We are now in the process of analyzing drinking outcomes
with the 3 modes for delivering MI. It seems quite feasible,though, to provide substance abuse counseling services torural populations via telephone or televideo links.
HEALTH CARE AS A CONTEXT FOR TREATING DRUG
ABUSE AND DEPENDENCE
Wilson M. Compton and Beverly Pringle
Because most individuals see a physician at least onceper year and because substance abuse and addiction arestrongly associated with many physical ailments, healthcare settings are promising places to screen for and addressthe myriad substance abuse and addictive problems. Yetthis promise is often unrealized and drug abuse services arerarely provided in primary care settings. Recent studiesshow that BIs may be useful for patients with drug-relatedproblems identified in general health settings and that pri-mary physicians can play an important role in improvingthe outcomes of persons treated in the specialty sector. Thepromise of these interventions is that they reach largenumbers of individuals with drug-related problems whodo not seek or respond to specialist treatment.
Key Issues for Treating Drug Abuse and Addiction in HealthCare Settings
Given the long-term relationship that many individualshave with medical providers, it is promising to considerhow general health care settings might be modified tomake them appropriate for treating drug abuse and addic-tion. Particularly as the treatment of addiction is viewed inthe context of an acute, relapsing condition, the role of thehealth care system should be emphasized (McLellan et al.,2000; Samet et al., 2002). Several issues will be importantas physicians attempt to realize this promise. The first key
298 MILLER ET AL.
PDF Create!
5 Trial
www.nuanc
e.com
issue is that specific drug abuse and addiction rates varydepending on the population being treated. For example,at present, methamphetamine may be encountered inmany health care settings in Hawaii and Southern Califor-nia but will be quite rare in New England. Even withinspecific regions, the type of practice will influence whatdrugs should be screened. Nearly all populations should bescreened routinely for alcohol and tobacco, but the situa-tion is less clear for illicit drugs. For example, marijuana isparticularly commonly abused by adolescents and youngadults and so screening for its use may be especially ap-propriate for adolescent pediatric practices. Screening rou-tinely for marijuana in a geriatric practice, on the otherhand, is less likely to be supported. Emergency rooms,with their high proportion of trauma and other at a riskpatients, are at a high risk for having an underlying addic-tion and so are a natural target for screening and interven-tion. Public clinics may be particularly important settingsas well, and a recent report from Bernstein et al. (2005)documents the potential utility of a brief peer-deliveredmotivational intervention in a general health clinic forreducing cocaine and heroin use at 3 and 6 months.The second key issue is that integrating general health
care into drug treatment can be useful for both addictiveand general health outcomes. Weisner et al. (2001) showedthat people in treatment for substance use disorders whoalso had significant substance-related medical conditionsshowed improved outcomes when treated in a system thatintegrated primary medical care with addiction treatment.Futher, Samet et al. (2003) have shown that assertive link-age with primary care during a detoxification treatmentcan be a successful engagement process. More recently,they showed that provision of primary medical care is as-sociated with improved addiction outcomes (Saitz et al.,2005). Thus, general medical care can play an essential rolein improving drug addiction outcomes.The third key issue is that screening instruments that are
brief and easy to administer have been developed foradults and adolescents in general medical settings. For ex-ample, the World Health Organization alcohol, smokingand substance involvement screen test (ASSIST) hasshown good reliability and feasibility (WHO ASSISTWorking Group, 2002). Current work on this instrumentis designed to determine its validity and clinical utility inmultiple settings. For adolescents, a promising develop-ment is the 6-item CRAFFT screening test for alcohol-related and drug-related problems and disorders. Thisbrief instrument has demonstrated good sensitivity, speci-ficity, and validity in adolescent clinic patients (Knight etal., 2002). These types of instruments may be useful inassuring consistent and adequate screening for addictiveproblems in health care settings. However, it is importantto recognize that integrating addiction screening andintervention/referral is less well established for illicit drugsthan for alcohol and tobacco. Major questions to beresearched relate to determining which drugs should
be screened and which patient populations should bescreened. In addition, a desire for thoroughness will haveto be balanced against the realistic time constraints in mosthealth care settings.
Conclusions
Developing efficacious BIs for addictive disorders is anessential but insufficient step toward improving publichealth. The problem is that development of an effectiveintervention does not ensure its widespread adoption. Forexample, in the case of smoking, implementation of inter-ventions in primary care settings that are successful inreducing smoking decreases substantially once activeresearch support is withdrawn (Hollis, 2005). The key dif-ficulties may lie in the areas of lack of clinician time, poorreimbursement for behavioral intervention services, inad-equate training, and inadequate staffing. New research isneeded that can address these issues and test viable sys-tematic changes in financing, organization, and manage-ment of clinical practice. For instance, given the limitedtime for screening and intervention for all behavioral con-ditions, how might screening tools address the full range ofsubstance and nonsubstance behavioral conditions. Atpresent, each set of conditions (mood, anxiety, alcohol,tobacco, and drug related) is generally addressed with aseparate screening questionnaire/technique. How canthese be efficiently packaged and structured so that theywill be useful and used by large numbers of clinicians? Thelessons from attempts to include tobacco, alcohol, andmental health interventions in general health care settingsindicate that even with documented efficacy, adoption,and widespread dissemination is not certain. This area ofdissemination and adoption research remains to be tackledwith clear and well-designed trials and will be a majorchallenge for drug abuse services researchers.
INTERVENTIONS FOR HEAVY DRINKING IN HEALTH
CARE SETTINGS: BARRIERS AND STRATEGIES
Mark L. Willenbring
The efficacy of screening and BI for heavy drinking inprimary care settings has been clearly established (Whit-lock et al., 2004). The World Health Organization, whichhas conducted a lengthy program of research and devel-opment in this area (Babor and Grant, 1992; WHO AS-SIST Working Group, 2002), has provided a core set ofstudies, instruments, and resources for this effort. Manycountries in the European Union have collaborated in re-cent years to work with governments to implement ASBI.Health care organizations in the United States, includingthe Veterans Health Administration, have devoted timeand effort to this endeavor as well. The United StatesPrevention Task Force has added ASBI to its ‘‘B’’ list ofevidence-based interventions (Whitlock et al., 2004).
299HEALTH CARE SETTINGS
PDF Create!
5 Trial
www.nuanc
e.com
In spite of this progress, screening rates in primary caresettings remain remarkably low. Even though efficacy isclear, PCP have mostly failed to implement ASBI in eve-ryday clinical practice. What I wish to do here is to brieflyoutline several key barriers and to suggest some possiblestrategies to move the effort ahead. Except where noted,my own observations and experience serve as the basis forthe thoughts outlined below.
1. Time and money. As anyone who has ever asked aboutadding a new task to someone’s workload knows, noone has enough time, and there is never enough money(usually expressed in terms of staffing levels, equip-ment, or administrative support). These 2 barriers arementioned so often that it is almost meaningless to citethem; yet we must address them. Physicians and nursesare indeed spending less time with patients, and therehas been constant budget pressure within health caresystems. However, another meaning to this reportedbarrier is that other tasks take priority, and there is notenough time and money (staff, equipment) to completeall desirable tasks. This problem has been amplified inrecent years by the proliferation of prevention recom-mendations in the absence of guidance regarding prior-ities or implementation. I will address these issues morewhen I discuss physician socialization and role.
2. Lack of incentive. In most cases, ASBI cannot currentlybe coded as a specific procedure for which additional re-muneration is provided. In other cases, physicians aresalaried, but their performance is based on other factors,such as patient panel size. However, other preventionprocedures such as breast examination and cholesteroltesting are conducted much more regularly even if notuniversally in the absence of such incentives. One mustbe cautious about assuming that providing an externalincentive would increase ASBI. Financial incentiveshave mixed evidence for efficacy in changing physicianbehavior (Town et al., 2004, 2005).
3. Lack of skill or training. It is true that physicians arenot well trained in ASBI, but it has not been demon-strated that training of physicians increases ASBI. Per-haps developing skills and knowledge are necessary butnot sufficient for implementing ASBI.
4. Concern about patient reaction. Physicians frequentlyexpress concern that patients will become angry or de-fensive if asked about their alcohol use and may evenchoose a different doctor because of it. This concern maybe linked to a lack of self-efficacy, as physicians may nothave confidence in their ability to skillfully manage suchreactions when they do occur.
5. Physician socialization and role. Physicians are social-ized primarily to diagnose and treat disease, not to pro-mote health. Socialization involves more than acquisitionof knowledge and skills; it is an assumption of an iden-tity and values, involving deeply held beliefs, many ofwhich may be unconscious. Television dramas and
popular press articles about doctors seldom focus onthe humdrum activities of preventive medicine andpatient education. Instead, heroic effort directed againstacute illness (primarily involving hospitals) providesthe backdrop for the inevitable soap opera involvingprincipal characters. Indeed, complex diagnosis anddisease management provide some of the most intel-lectually stimulating challenges of medical practice.Relieving suffering, improving function, and curingdisease are particularly gratifying activities as well.Most patient visits are initiated because of a specifichealth concern, and patients are likely to be impatientwith doctors who divert attention to attend to preven-tion not related to current symptoms. Finally, payingphysicians to conduct primary and secondary preven-tion, along with other behavioral interventions such asthose focused on disease management or treatmentadherence, is probably not cost-effective. It is likelythat a health educator or nurse could provide the sameservice with equal or greater efficacy at a less cost.
6. Failure to address the substance-dependent medical pa-tient. The prevalent model for treating nonopioid sub-stance dependence (rehabilitation) consists of relativelyintensive, time-limited psychosocial therapy for patientswilling to endorse a goal of long-term abstinence. ASBIaddresses nondependent patients, and rehabilitationaddresses a minority of dependent patients. However,the US addiction treatment system is in crisis (McLellanand Meyers, 2004). Many patients have no reasonableaccess to treatment. Furthermore, addiction treatmentpersonnel have provided very little assistance to physi-cians in managing substance-dependent patients whohave coexisting medical or serious mental disorders, butwho are either unwilling to commit to a goal of absti-nence or unable to maintain long-term abstinence.Thus, physicians may be concerned that screening willidentify many dependent patients who have no accessto treatment, who refuse referral, or who have notresponded. Although the focus of ASBI is on the non-dependent patient, screening identifies all heavy drink-ers, including dependent drinkers.
7. Discrimination and stigma. Heavy drinking and alco-hol use disorders are still heavily stigmatized condi-tions, which discourages ASBI and also discouragesaccurate medical records. Anticraving medicationsmay not be covered by insurance plans in spite of evi-dence supporting efficacy, but the public relations costto third-party payers is minimal because of a lack ofactive advocacy.
What strategies might we use to increase ASBI in med-ical settings?
1. Decide what role physicians (vs other health profes-sionals) should play in ASBI. As physicians are unable
300 MILLER ET AL.
PDF Create!
5 Trial
www.nuanc
e.com
to accomplish all recommended prevention tasks andmay prefer to diagnose and manage disease, otherhealth professionals may be better suited to ASBI.Nurses, behavioral health care specialists, and healtheducators are likely to do a better job at a less cost. Atthe same time, the support of physicians is likely to beimportant to patients. Additional research is needed totest various models of implementing such a comple-mentary or team approach.
2. Provide incentives. Physicians and health care organi-zations are unlikely to add services that are not reim-bursed even if they are a good idea. Alcohol screeningand brief intervention should be a reimbursable proce-dure, even if done by a nonphysician. Better yet, dem-onstrating a reduction in heavy drinking in a clinicpopulation or physician patient panel could be reward-ed with bonuses by third-party payers.
3. Consider targeted as opposed to universal screening.Rather than expecting ASBI on each visit (a practicethat may not be cost-effective anyway), consider an ap-proach targeting situations where heavy drinking ismore likely or of greater concern. Examples includetrauma (especially motor vehicle crashes), smokers, in-somnia, liver disease, pregnancy, or when prescribingmedications.
4. Help physicians better manage the heavy drinkers theyalready know about. Every physician has patients whoare known to be heavy drinkers. It is likely most ofthem are alcohol dependent, which ASBI does nottarget. Physicians may be concerned, however, thatscreening may also identify dependent patients they donot know how to manage. Disease managementapproaches have demonstrated efficacy in people whodrink heavily who have alcohol-related medical disor-ders (Lieber et al., 2003; Miller et al., 2001; Willenbringand Olson, 1999a, 1999b). Providing physicians witheffective methods for managing alcohol-dependentpatients, especially those who have co-existing medicaland/or psychiatric disorders, would not only improvethe management of these patients but may also makephysicians more receptive to and interested in ASBI.Research evaluating the integration of disease manage-ment strategies for alcohol dependence with those forother chronic disorders is needed.
5. Provide information to consumers. Consumers are veryunlikely to have current information about treatmentfor alcohol dependence and are unlikely to inquire aboutit from their physicians. Providing products for consum-ers concerning evidence-based treatment may be an ef-fective strategy to increase consumer demand for them.
REFERENCES
Babor TF, Grant M (1992) Project on Identification andManagement of
Alcohol-Related Problems: Report on Phase II: A Randomized Clin-ical Trial of Brief Interventions in Primary Health Care. World Health
Organization, Geneva, Switzerland.
Beich A, Gannick D,Malterud K (2002) Screening and brief intervention
for excessive alcohol use: qualitative interview study of the experiencesof general practitioners. Br Med J 325:870–874.
