Linking substance abuse services with general medical care: integrated, brief interventions with hospitalized patients

Linking substance abuse services with general medical care: integrated, brief interventions with hospitalized patients

Christopher W. Dunn


Brief Interventions

Estimates of the prevalence of harmful substance abuse among general hospital populations vary with the medical problems addressed, but they are at least twice that of the general population (1, 2). Yet a substance abuse task force at Harborview Medical Center in Seattle, where this study was conducted, found that except for the emergency department, which routinely screens trauma patients for substance abuse, nearly 50% of substance-abusing nontrauma patients elsewhere in the hospital were still not routinely identified and counseled (3). This paper presents the short-term results of implementing on-site alcohol or drug (AOD) brief interventions across several medical, surgical, and psychiatric services in this hospital.

Researchers have evaluated several successful efforts to combine medical care and substance abuse treatment in medical settings (4, 5). Because of time limitations in medical settings, most of these efforts provided patients with rapid substance abuse assessment and brief interventions (5). A brief intervention usually involves a short counseling session during which the counselor or medical provider assesses the patient’s substance use, gives the patient feedback on the assessment results, and then negotiates a plan of action with the patient to change his or her behavior.

The empirical support for brief substance abuse interventions is excellent. For example, a 1995 meta-analysis of 32 alcohol treatment modalities found that brief motivational counseling ranks near the top in four categories: 1) total amount of research performed to investigate the modality, 2) methodological quality of that research, 3) number of studies showing positive outcomes, and 4) cost-effectiveness (6). Additionally, Bien’s review of brief interventions for alcohol problems reported that across different medical settings and countries, seven of eight randomized trails yielded significant reductions in alcohol use and/or related drinking problems. Two other randomized studies found that patients receiving versus not receiving brief counseling participated more fully in treatment and consumed less alcohol 3 months following that treatment (7, 8).

In addition to their brevity, cost-effectiveness, and positive outcomes, another reason for initiating brief interventions in general medical settings is their timeliness. Gentilello (9) has shown that alcohol-related trauma may create a teachable moment during which patients may be particularly receptive to counseling. Similarly, hospitalized general medical patients might come to see a link between their substance abuse and their health problems after brief, opportunistic interventions.

Brief substance abuse interventions can potentially reduce health-care costs. Certainly, many patients need intensive substance abuse treatment such as inpatient or outpatient programs combined with 12-step meetings. A brief intervention would not suffice. In these cases, brief interventions serve to help the patient become motivated to seek this intensive treatment sooner than he or she might without help. The recent CALDATA study on the cost-effectiveness of substance abuse treatment documented pretreatment to posttreatment declines in hospitalizations for physical health, drug overdose, and mental health of 36%, 58%, and 44%, respectively. Also, the mean number of hospital days declined by 25% after treatment (10). Given that for some patients, brief substance abuse interventions yield similar effect sizes as more intensive treatment (11), it is reasonable to expect medical cost reductions from brief interventions.

Motivation for Change

A simplified version of Prochaska’s and DiClemente’s transtheoretical model of change, often associated with motivational counseling, construes change as occurring in stages: precontemplation (not seeing a problem or considering change), contemplation (sees the problem but is ambivalent about change), and action (ready for action very soon or already in action to change) (12). Accordingly, motivation is seen as a matter of degree, that is, the probability that a patient will take action to change a destructive behavior (13). Under this model, motivation is operationalized by assessing which stage of change readiness the patient is in currently.

Whereas more traditional substance abuse treatment typically targets how the patient is going to take action to change (go to AA, go to aftercare, identify relapse triggers and use relapse prevention tools), brief motivational interventions focus on why the patient might consider change (to feel better about self, to be a better father, to save money), and if so, what the patient wants to change (quit which drugs? for how long? improve your health?). This leads to the patient and counselor negotiating a change plan when the patient is ready to do so. For example, this plan might range anywhere from doing absolutely nothing differently to going for complete abstinence by means of inpatient, outpatient, aftercare, and 12-step meetings.

Brief interventions for substance abuse have been tested more often in primary health-care (14, 15) or emergency/trauma (9, 16) populations than in general medical settings. Most studies that have tested substance abuse screening, intervention, and referral protocols have used designated intervention teams that “floated” throughout the hospital, responding to consult requests. Recently, a study used a floating consult team to intervene with general medical patients in a general hospital/trauma center. This protocol was feasible and cost-effective, although follow-up was not performed to assess patient compliance with treatment referrals or substance abuse outcomes (17). The current study used fully integrated (on-site) brief interventions (IBI) with the general medical/surgical/psychiatric patients in a county hospital. We anticipated that patients might resist less if they perceived the on-site services as being a standard part of their medical care rather than as an unrelated outside service. Also, we expected that ward staff, by virtue of their daily contact with on-site chemical dependency professionals, might come to know and trust them better than “floaters,” and might refer to them more often.


Harborview Medical Center includes a 400-bed county general hospital/trauma center staffed by the University of Washington, which serves mostly indigent, emergently admitted medical, psychiatric, and surgical trauma patients. A pilot program was implemented to provide IBI within the structure of general hospital services. IBI refers to on-site, on-demand substance abuse consultation and brief intervention by alcohol or drug (AOD) specialists. These onsite specialists provided evaluations, feedback, and treatment recommendations for patients in the following areas: drugs of choice, presence and severity of alcohol or drug abuse/dependence, previous or current treatment participation, and motivation toward recovery (stage of change readiness). Note that in this paper the word referral has two uses. It means a medical staff member’s initial referral of a patient to the on-site AOD specialist. It also means the subsequent referral for substance abuse treatment that often occurs after the specialist assesses the patient, gives feedback, and makes recommendations. The following research questions were formulated:

1. Which medical services would use IBI services the most? 2. Would medical services make appropriate referrals to the IBI specialists, or would there be many false positives? 3. What substance abuse profiles would emerge for patients referred to the IBI services: What drugs of choice? What substance diagnoses? How severe might their problems be? 4. What would be the treatment histories and current levels of motivation for treatment among referred patients? 5. To what degree would patients comply with referrals?

Assessment and Intervention Methodology

Implementing the IBI model made no changes in how the hospital handled the commonly encountered problems in substance abusers such as the medical management of withdrawal states or the formulation of differential psychiatric diagnoses. As usual, these problems were triaged by medical or psychiatric staff. Implementing the IBI model involved assigning four state-certified CD counselors to work full-time on the following hospital services: Psychiatry (inpatient), Ob/Gyn (outpatient), Surgery (inpatient), and Adult Medicine (outpatient). Medical, nursing, and social work staff on these services were made aware of the IBI services and encouraged to refer patients whenever clinical impressions or patients charts suggested possible drug/alcohol problems. Only CD problems were referred to CD staff. All other psychiatric problems identified by staff–or by CD counselors once they interviewed the patient–were triaged by psychiatric consultants. Once referred, patients were consented for the study and then underwent IBI (assessment and brief counseling).

CD counselors performed assessment interviews of about 45 minutes, during which they completed structured demographic and substance abuse data forms. All Likert-scale items had specific anchor points. The CD counselors made DSM-IIIR diagnoses of substance abuse or dependence using data gained from interview questions formulated by the counselors to assess how many DSM-IIIR criteria were met by each patient. These questions were not scripted; instead they were asked conversationally, within the context of the assessment interview. Patients’ motivation for change was scored using a single 5-point Likert rating summarizing the counselor’s clinical impression of the patient as: a) not motivated at all (1-2), b) partially motivated but still somewhat ambivalent (3-4), or c) fully motivated for recovery (5). These three categories correspond respectively to three stages of change readiness in Prochaska’s and DiClemente’s transtheoretical model: precontemplation, contemplation, and action (14, 16).

Following these interviews, counselors performed brief counseling with patients by giving them feedback about diagnoses and problem severity and then negotiating appropriate referrals or recommendations. Referrals for AOD services were made on the basis of how realistic it was for a patient to pursue a given referral source, rather than on problem severity. Specifically, the counselor considered the patient’s psychiatric status (severe dual diagnoses ruled out most community treatment programs), mobility (severely injured or ill patients were unable to enter outpatient treatment immediately), geographic location, financial resources, and motivation for treatment.

Counselors factored patients’ motivational stages into their referral or recommendations as follows. Consider the relatively common example of a patient in precontemplation who was assessed by the counselor as having a problem severe enough to warrant outpatient treatment, yet that patient lacked the readiness to voluntarily take action by entering treatment on the counselor’s recommendation. In such a case, the counselor would state his or her professional opinion-that treatment would be the best course of action for that patient to take when the patient becomes ready in the future to go into action to change his drinking/drug use. Rather than cause an argument with a resistant client, the counselor might then use the remaining time to provide that patient with information about the negative consequences of his substance abuse, in order to raise doubt about the harmful behavior, and move the patient closer to contemplation and action. In this way, unrealistic referrals were avoided, so as to prevent subsequent failure experiences and to make the referral process seem more credible and of value to patients. At the time of referral, patients were told of problems of treatment accessibility. With the more motivated patients (action stage), counselors helped patients prepare for dealing with barriers to getting into publicly funded treatment.

Based on patients’ acceptance of these diagnoses and referrals, a plan of action was negotiated along a continuum ranging from taking no action whatsoever to immediately entering professional treatment plus participating in 12-step meetings. In the case of patients less ready for taking action, counselors tried to negotiate a verbal agreement with the patient that he or she would step up the plan, should it fail. For example, a patient wanting to quit drinking but insisting on doing it without outside help might verbally agree to attend AA or seek treatment if he or she should start drinking again. Follow-up phone calls were attempted by counselors 1-2 weeks later. These follow-up assessments rated stage of change readiness and any action taken by patients toward recovery.


Three hundred sixty-three patients consecutively referred by hospital staff were evaluated by CD counselors over a 3-month period. Table 1 summarizes patient demographics and the hospital services from which these patients were referred. Ethnic and gender distributions were consistent with Harborview’s overall population. Seventy-four percent of all referrals during this time came from Psychiatry. Only 5% of the sample (N = 17) were determined not to have a diagnosis of substance abuse or dependence. Nine percent of the sample (N = 34) were diagnosed as substance abusing, and 83% (N = 301) were diagnosed with substance dependence ranging from mild (met minimal DSM-III criteria) to severe (met most or all criteria). Of those diagnosed with substance dependence, 63% were severely dependent. For 11 patients, counselors were unable to form accurate diagnoses because they were given incomplete responses. Over 50% of the sample reported multiple substance use (see Table 2). Although only 292 of 363 patients answered the question about IV drug use, 18% of those patients reported using IV drugs at least once per month. Alcohol was overwhelmingly reported to be the drug of choice (66% versus 13% and 7% for cocaine and opiates, respectively).

Table 1. Referred Patient Demographics and Hospital Referral Service (N = 363)

Male Female Inpatient Outpatient Other(a)

238 (66%) 125 (34%) 301 (83%) 62 (17%) 10 (3%)

Caucasian Afr.-Amer. Hispanic Native Amer. Asian

252 (69%) 72 (20%) 12 (3%) 13 (4%) 4 (1%)

Psychiatry Ob/Gyn surgery Adult med. and Other(a)

267 (74%) 24 (7%) 20 (6%) 52 (13%)

(a) “Other” included N = 35 consults performed for nearby services such as Ob/Gyn or Pediatrics.

Table 2. Drugs of Choice for Referred Patients(a)

1st Choice 2nd Choice 3rd Choice Total

Alcohol 239 (66%) 49 12 300

Cocaine 47 (13%) 50 8 105

Opiates 27 (7%) 8 3 38

Cannabis 26 (7%) 17 11 54

Amphetamines 3 (1%) 7 4 14

Hallucinogens 6 (2%) 5 3 14

Prescription 5 (2%) 4 0 9

Other 10 (3%) 3 3 16

Total 363 (100%) 143 44 550

(a) Polydrug users N = 187 (52%). Self-reported IV drug users N = 64 (17%). Percentages computed using N = 363.

Nearly two-thirds of the sample (62%) had been exposed to some form of professional treatment in the past, and over half had prior Alcoholics, Narcotics, or Cocaine Anonymous (AA, NA, or CA) experience. Seventy-nine percent had neither current professional treatment nor AA, NA, or CA involvement. Eighty-three percent of patients needed to qualify for public funding in order to get treatment. The remainder of the sample had private funding (15%) or self-pay (2%) capabilities.

Patients’ motivation (stage of change readiness) was assessed both during the evaluations and during brief follow-up telephone contacts 1-2 weeks later. At the time of baseline assessments in the hospital, 21% were in the precontemplation stage, 54% were in contemplation, and 20% were in action. For follow-up, 187 of 363 patients were contacted, and the distribution of stages of change readiness of those contacted was as follows: precontemplation (10%), contemplation (51%), action (38%). Using an unlikely worst-case scenario in which all noncontacted subjects are considered to be in precontemplation, 20% of the full sample were in the action stage at follow-up versus 21% at baseline. At follow-up, 125 of 187 (66%) patients contacted reported being involved in professional treatment or AA,NA, or CA meetings, as compared to 75 of 352 at baseline (22%). When calculated as a worst-case scenario (e.g., none of the 165 patients not contacted for follow-up were involved in treatment), 35% of the original 352 patients reported involvement in some kind of treatment (including AA, NA, or CA meetings) at follow-up, versus 21% at baseline. The low percentage of follow-up contacts likely reflects both the interventionists’ work loads and the public/indigent, often homeless, nature of the population.


Data for this study were collected not by research staff in a research setting but by AOD clinicians in a clinical setting that limited the quantity and quality of accessible data. Clinicians used nonvalidated instruments, although the structured rating scales had well-described anchor points. Because of financial constraints, interrater reliability checks were not performed on counselors’ stage of change ratings. These data likely reflect the kind of clinical information that can realistically be gathered in medical settings which seldom afford adequate time or confidentiality. For this reason and because data on total admissions for general medical patients were not collected, prevalence estimates of substance abuse/dependence among this sample are not available.

Most referrals to the IBI services in this project came from the hospital’s psychiatric wards. Compared to Psychiatry, the other three hospital services used IBI much less. Whereas the other services in this hospital had only recently acquired IBI services, Psychiatry had a developed AOD intervention track, its dual disorders program. The presence of a dual disorders program and the fact that the training backgrounds of psychiatric staff impart more AOD awareness and skills than the training backgrounds of general medical staff could account for the higher IBI utilization rate by Psychiatry. Also, there is likely a higher base rate of substance abuse among psychiatry patients than among general medical patients. Our data suggest that in-service training of non-Psychiatry hospital staff might increase awareness of substance abuse and willingness to use IBI services. Indeed, the act of referring a patient to the IBI service is itself a “mini-intervention” demanding specific skills that can be taught to medical staff members.

How accurate were the clinical impression of hospital staff regarding substance abuse/dependence? Clinical impressions on the part of hospital staff appeared to be accurate, given that only 5% of referred patients were found to have no diagnoses of substance abuse/dependence. But in light of the large proportion of patients diagnosed with severe dependency, it seems possible that staff members may have failed to detect and refer some patients with mild or moderate substance problems. Since brief interventions are thought to be particularly effective with mild severity cases (5), hospitals should consider implementing standardized screening systems that are sensitive enough to identify cases of mild and moderate severity.

What were the treatment histories and levels of motivation among referred patients? Almost two-thirds (62%) of respondents had prior professional treatment and/or AA, NA, or CA experience. However, 76% were not currently involved in either treatment or meetings. The fact that 54% were classified as ambivalent (would at least consider treatment) and 20% as fully motivated for recovery suggests an unmet need for treatment among these patients. Accessing publicly funded treatment, the most realistic alternative for most of our patients, can take several weeks and requires a high level of motivation. Unfortunately, patients who are ambivalent–in the contemplation stage–might lose their resolve in the face of this red tape, even though treatment might benefit some ambivalent patients once they get there (12).

In this study, motivation was scored using a single scale item consisting of 5 points defining readiness to change. Because of its brevity, this method is clinically preferable to other, more time-consuming means, such as using the Readiness to Change Questionnaire. Rollnick used this instrument to assess readiness to change among a similar population of nontreatment-seeking problem drinkers in a general medical ward (18). The distribution of stages of change readiness was similar to the distribution found in the current study, using a single scale item (17). It would be helpful for future research to develop and validate single scale items, or otherwise very brief readiness to change instruments for clinical use in medial settings. Assessing a patient’s stage of change readiness is essential for guiding the clinician in choosing an appropriate counseling change strategy (12).

Due to our low follow-up rate (52%) and lack of control group, we are uncertain whether or not patients’ motivation to change increased in the short run as a result of the brief interventions performed in this study. Because those patients who were reached for follow-up might be higher functioning and more motivated than those lost to follow-up, little may be concluded about the motivational effects of the brief interventions on the full sample. Many of our patients who reported being involved in treatment at follow-up had done so in Harborview’s inpatient or outpatient dual disorders program, thus bypassing public treatment delays. Brief interventions, whether integrated on site or floating, might be most effective at placing patients in treatment if the treatment itself is available on-site, on demand.

IBIs could potentially benefit precontemplators and contemplators. Such patients, who are not yet in the action stage, might continue to drink or use and thus get discharged early from more traditional abstinence-oriented community addiction treatment. Such a treatment outcome can consume scarce community resources and discourage precontemplators or contemplators from seeking treatment again when they have become more motivated for change. IBI can also be seen as a bridge to more intensive treatment if needed, as a way to reduce harm among precontemplators and contemplators not yet interested in abstinence-oriented action. As their motivation progresses and they move closer to the action stage, they should be more likely to commit to the goal of total abstinence (13, 19).

If a substantial number of substance-misusing general medical patients could benefit from brief interventions (5), it makes sense to implement on-site brief interventions in hospitals as described in this study. This approach could boost the quality of medical care while providing a context for counselors to work with precontemplators and contemplators, who might be mismatched for community abstinence-oriented programs (19). From a treatment matching perspective, totally abstinence-focused treatment may be most appropriate for patients ready for action, whereas brief interventions might be appropriate for people in precontemplation or contemplation by helping them prepare for change as quickly as possible, possibly reducing harm in the process. We are not advocating that precontemplators and contemplators be kept out of abstinence-oriented treatment. However, if a patient is willing to undergo a brief intervention in the hospital but unmotivated or unable to enter intensive community treatment immediately, a brief intervention is better than nothing.

The Institute of Medicine has reported that most studies on the duration of treatment have found no overall advantage for longer-term versus shorter-term treatment (20). Some brief interventions shown to be effective have involved as little as 5 minutes of advice given to patients by doctors or nurses. In some cases, these brief interventions resulted in 6-month and 12-month reductions in drinking and/or drinking-related problems that were comparable in effect to more intensive but less accessible forms of AOD treatment (5, 11). Future research should evaluate the cost-effectiveness of integrated brief interventions among general medical patients who screen positive for substance abuse problems. Just as user-friendly, brief interventions can help substance abusers move from the denial of precontemplation to the commitment stage of action, integrated brief interventions may be a user-friendly technology to help medical centers become more committed to changing how AOD problems are addressed.

Table 3. Baseline Dependency, Treatment Involvement, and Motivation Level of Patients(a)


Category Low rating rating

Degree of AOD None: 17 (5%) Abuse: 34 (9%)

involvement (no DSM-III (DSM-III

diagnosis) Abuse only)

Current AOD No meetings or Monthly

treatment treatment: meetings or

or AA/NA/CA 277 (76%) treatment:

22 (6%)

Previous AOD Never: 127 One or more previous

treatment (36%) treatment attempts: 224 (62%)

Previous Never or only 10-100 meetings:

AA/NA/CA one meeting: 93 (26%)

161 (44%)

Acceptance Refused or Ambivalent:

of referral were evasive: 118 (33%)

132 (36%)

Motivation None: Ambivalent:

level 76 (21%) 196 (54%)

Category High or severe rating

Degree of AOD Mild-Severe DSM-III

involvement Dependency: 301 (83%)

Current AOD Meetings or treatment, weekly

treatment or more often: 53 (15%)


Previous AOD One or more previous

treatment treatment attempts: 224 (62%)

Previous More than 100 previous meetings

AA/NA/CA 97 (27%)

Acceptance Fully accepted referral: 99 (27%)

of referral

Motivation Fully motivated for recovery:

level 72 (20%)

(a) Percentages do not total 100% due to incomplete responses to some questions. Percentages are computed using N = 363.


This work was supported in part by grant #5 R01 AA09045 from the National Institutes of Health.


(1.) Barrison, I. G., Viola, L., Mumford, B. A., Murray, R. M., et al., Detecting excessive drinking among admissions to a general hospital, Health Trends 14:80 (1982)

(2.) McIntosh, D., Alcohol-related disabilities in general hospital patients: A critical assessment of the evidence, Int. J. Addict. 17(4):609 (1982).

(3.) Harborview Medical Center Substance Abuse Task Force, Community Needs Assessment, Seattle, Washington, 1995.

(4.) Levin, S. M., Trumble, J. G., Edmunds, M., et al. Perspectives on linkage of primary health care and substance abuse treatment, J. Addict. Dis. 12:1 (1993).

(5.) Bien, T. H., Miller, W. R., and Tonigan, J. S., Brief interventions for alcohol problems: A review, Addiction 88:315 (1993).

(6.) Hester: R. K., and Miller, W. R., Handbook of Alcoholism Treatment Approaches: Effective Alternatives, Allyn & Bacon, Boston, 1995.

(7.) Brown, J. M, and Miller, W. R., Impact of motivational interviewing on participation and outcome in residential alcoholism treatment, Psychol. Addict. Behav. 7:211 (1993).

(8.) Bien, T. H., Miller, W. R., and Boroughs, J. M., Motivational interviewing with alcohol outpatients, Behav. Cognitive Psychother. 21:347 (1993).

(9.) Gentilello, L. M., Duggan, P., Drummond D., et al., Major injury as a unique opportunity to initiate treatment in the alcoholic, Am. J. Surgery 156:558 (1988).

(10.) Gerstein, D. R., Johnson, R. A., Harwood, H. J., et al. The California Drug and Alcohol Treatment Assessment, California Department of Alcohol and Drug Programs, Sacramento, 1994.

(11.) Miller, W. R., and Hester, R. K., Inpatient alcoholism treatment: Who benefits? Am. Psychol. 41:794 (1986).

(12.) Prochaska, J. O., DiClemente, C. C., and Norcross, J. C., In search of how people change: Applications to addictive behaviors, Am. Psychol. 47:1102 (1992).

(13.) Miller, W. M., and Rollnick, S., Motivational Interviewing: Preparing People to Change Addictive Behaviors, Guilford, New York, 1991.

(14.) Senft, R. A., Polen, M. R, Freeborn, D. K., et al., Drinking Patterns and Health: A Randomized Trial of Screening and Brief Intervention in a Primary Care Setting, Center for Health Research, Kaiser Permanente, Portland, 1995.

(15.) Babor, T. F., de la Fuente, J. R., Saunders, J., et al., The Alcohol Use Disorders Identification Test: Guidelines for Use in the Primary Health Care, WHO Publication No. 92.4. World Health Organization, Geneva, Switzerland, 1992.

(16.) Fuller, M. G., Diamond, D. L., Jordan, M. L., et al., The role of a substance abuse consultation team in a trauma center, J. Alcohol Stud. 56:267 (1995).

(17.) Fuller, M. G., and Jordan, M. L., The substance abuse consultation team: Addressing the problem of hospitalized substance abusers, Gen. Hosp. Psychiatry 16:73 (1994).

(18.) Rollnick, S., Heather, N., Gold, R., et al., Development of a short “Readiness to Change Questionnaire” for use in brief, opportunistic interventions among excessive drinkers, Br. J. Addict. 87:743 (1992).

(19.) DiClemente, C. C., Carbonari, J. P., and Velasquez, M. M., Alcoholism treatment mismatching from a process of change perspective, Drug Alcohol Abuse Rev. 3:115 (1992).

(20.) Institute of Medicine, Broadening the Base of Treatment for Alcohol Problems, National Academy Press, Washington, D.C., 1990.

Richard Ries, M.D. To whom correspondence should be addressed at Harborview Medical Center, University of Washington, Seattle, Washington 98104-2499. E-mail: Telephone: (206) 731-3423.

COPYRIGHT 1997 Taylor & Francis Ltd.

COPYRIGHT 2004 Gale Group