Screening and Brief Intervention in the Emergency Department

Gail D’Onofrio, M.D., M.S., and Linda C. Degutis, Dr.P.H.

Many patients visiting hospital emergency departments (EDs) or admitted to trauma centers have alcohol problems. Therefore, it is plausible that all ED and trauma patients should be screened for unhealthy alcohol use so that optimal care can be provided and treatment initiated, if necessary, for these patients. In addition, brief interventions offered directly in the ED or trauma unit could be useful for many patients. Some studies have found such interventions to be feasible and effective in this setting. However, all efforts in this regard must take into consideration the specific challenges associated with screening and intervention in EDs, such as time constraints, ethical and legal issues, and concerns regarding insurance coverage. Innovative approaches to screening may address at least some of these problems, although more research is needed to determine how screening can be better incorporated and implemented in the ED setting. KEY WORDS: hazardous AOD (alcohol and other drug) use; harmful AOD use; alcohol abuse; emergency room; trauma center; drinking and driving; identification and screening; intervention (persuasion to treatment); brief intervention; counseling; motivational interviewing; barriers to treatment; literature review

any patients visiting hospital ing and brief interventions in this setting, (D’Onofrio and Degutis 2002). In emergency departments including ethical and legal barriers to addition, patients treated in EDs are M(EDs) exhibit unhealthy screening and intervention. In addition, 1.5 to 3 times more likely than those alcohol use (Saitz 2005), which encom- innovative approaches to screening and passes patterns of alcohol consumption intervention in the ED are presented, GAIL D’ONOFRIO, M.D., M.S., is an that put the drinker at risk for adverse and issues that need to be addressed in associate professor in the Department of consequences (known as at-risk drinking), future studies are discussed. Surgery, Section of Emergency Medicine, have led to alcohol-related problems but and interim chief of the Section of do not meet the criteria for an alcohol Emergency Medicine, Yale University School use disorder (known as problem drink- Prevalence of Unhealthy of Medicine, New Haven, Connecticut. ing), or meet the criteria for an alcohol Alcohol Use in ED Patients use disorder (i.e., alcohol abuse or alcohol LINDA C. DEGUTIS, DR.P.H., is an dependence). (For more information, A substantial portion of the estimated associate professor in the Department of see the textbox.) As a result, ED practi- 110 million ED visits in the United Surgery, Section of Emergency Medicine, tioners routinely care for patients with States each year are related to unhealthy and research director of the Section of adverse health effects associated with alcohol use. As many as 24 to 31 percent Emergency Medicine, Yale University alcohol consumption. This article exam- of all patients who are treated in an ED School of Medicine, New Haven, ines the prevalence of alcohol-related and as many as 50 percent of severely Connecticut. health problems among ED patients, injured trauma patients (i.e., patients reviews studies evaluating the effective- who require hospital admission, usually Funding support for the preparation of ness and feasibility of brief interventions to an intensive care unit, for treatment this article was provided by National in the ED setting, and points out partic- of acute injuries) have positive results Institute on Alcohol Abuse and Alcoholism ular challenges associated with screen- when screened for alcohol problems grant R01–AA–12417.

Vol. 28, No. 2, 2004/2005 63 treated at primary care clinics to report heavy drinking, adverse consequences Definitions of Unhealthy Alcohol Use of drinking (e.g., alcohol-related injuries, illnesses, and legal or social problems), The term “unhealthy alcohol use” refers to a spectrum of disorders ranging or having ever been treated for an alco­ from at-risk drinking to alcohol dependence. At-risk or hazardous drinking hol problem (Cherpitel 1999). implies that the person is drinking over the recommended limits and is The prevalence of alcohol use disor­ therefore vulnerable to illness, injury, or social/legal problems. These recom­ ders in ED patients was confirmed by a mended consumption limits are, for men, 2 standard drinks per drinking study conducted in seven representative occasion or 14 standard drinks per week, and, for women and people age EDs across Tennessee, in which patients 65 and over, 1 standard drink per drinking occasion or 7 drinks per week. were assessed to determine their need A standard drink is defined as 12 grams of pure alcohol, the amount con­ for alcohol and other drug (AOD) tained in approximately 12 oz of beer, 5 oz of wine, or 1.5 oz of distilled treatment (Rockett et al. 2003). The spirits. researchers reported that, based on the Once a person experiences an alcohol-related harmful event—an injury, assessment, as many as 27 percent of illness, or social/legal problem such as poor grades, an argument with parents, ED patients needed AOD treatment or a driving violation—he or she is classified as a harmful drinker. services; however, in only 1 percent of The far end of the spectrum includes alcohol abuse and alcohol depen­ the cases did the ED physicians docu­ dence as defined by the diagnostic criteria that have been established in the ment a diagnosis of AOD abuse in the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental patients’ medical files. Moreover, less Disorders, Fourth Edition (DSM–IV). than 10 percent of the patients deter­ mined to be in need of AOD treatment actually received any. Patients were more on Drug Abuse (SAMHSA 2003b), 3 entire spectrum of unhealthy alcohol likely to need AOD treatment services million people ages 16 to 20 had driven use and that they be offered brief inter­ if they were insured by Medicaid or under the influence of alcohol at least vention and referral to treatment services. Medicare, had come to the ED 2 or more once in the past year, including 600,000 hours after the onset of the illness or people ages 16 or 17. Motor vehicle injury (i.e., had delayed the ED visit), crashes are the number one cause of Effectiveness of Brief or had visited the ED more than once. death in people ages 1 to 35, and the Interventions in the Interestingly, no difference in treatment eighth leading cause of death overall ED Setting need existed between patients visiting (CDC 2004). In 2003 (the most recent the ED because of an injury (e.g., from data available), there were approximately ED practitioners are chronically pressed an alcohol-related car crash or fall) and 43,000 motor vehicle traffic fatalities for time, and resources often are limited patients visiting the ED for other illnesses. in the United States, according to the in this setting. Therefore, if ED practi­ Finally, treatment need was inversely National Highway and Traffic Safety tioners are to be encouraged to screen associated with age—that is, younger Administration (NHTSA), of which their patients for alcohol problems and patients were more likely to need AOD an estimated 18,000 (40 percent) were offer brief intervention if necessary, treatment than older patients. related to the use or abuse of alcohol particularly under the time constraints Several factors contribute to the fact (NHTSA 2005). Consequently, NHTSA they are facing, they first must be that younger ED patients may be more has made prevention of impaired driv­ convinced of the usefulness of these likely to have alcohol-related problems ing a major initiative and is working measures. Brief interventions are short that indicate a need for treatment. First, to encourage health care practitioners counseling sessions, ranging from 5 to younger people are usually healthy and (including those in EDs) to provide 60 minutes, performed by nonaddic­ are more likely than older people to be screening and brief intervention services. tion specialists. Including the concepts uninsured and to use the ED as their This initiative supports the Institute of motivational interviewing (MI) may usual source of care (McCarthy et al. of Medicine’s (1990) landmark report enhance the success of the intervention 2002).1 Second, young adults have the on broadening the base of AOD abuse in changing patients’ behavior. The highest prevalence of binge2 and haz­ treatment, which recommends that principles of MI, developed by Miller ardous drinking in the United States patients in all medical settings should and Rollnick (1991), are encapsulated (SAMHSA 2003a), which can easily be screened for problems along the in the FRAMES acronym (feedback, escalate to drinking patterns that require responsibility, advice, menu or choice, 1The various studies define “older” and “younger” differ­ intervention. ently. Most commonly, “younger people” are those age 25 empathy, and self-efficacy). The goal Third, particularly in younger people, and younger. of the brief intervention is to assist these drinking patterns often occur in patients who exhibit less severe patterns 2Binge drinking as it is used here is defined as consump­ conjunction with driving. According to tion of five or more drinks on one drinking occasion; it also of unhealthy alcohol use (i.e., at-risk the 2001 National Household Survey is called heavy episodic drinking. drinking and problem drinking) to

64 Alcohol Research & Health Screening and BI in the Emergency Department

reduce their alcohol consumption to teens in an urgent care setting” and Participants were randomly assigned to low-risk levels, thereby reducing the risk included a handout on avoiding drink­ one of three groups: of illness or injury. For those patients ing and driving as well as a list of local who are alcohol dependent, the goal treatment agencies. The 94 participants •Standard care (SC) (N = 188). may be abstinence and/or acceptance were recruited for the study because they of a referral to a specialized treatment had positive blood alcohol concentrations •Abrief intervention (BI) consisting program. (BACs) or an alcohol-related injury (i.e., of a 40- to 60-minute session provided Chafetz and colleagues (1962) pub­ reported drinking prior to the injury by non-ED staff (i.e., a social worker lished the first report on what could be that required treatment). Followup or graduate student) (N = 182). considered a brief intervention in the assessments were conducted by phone ED (i.e., a referral to a specialized after 3 months and through face-to-face • BI with a booster (BIB) that entailed alcoholism treatment clinic as well as interviews after 6 months. This study a scheduled return visit 7 to 10 days information on strategies to assist with found that both groups of patients after the initial BI (N = 169). social issues, etc.). Their study demon­ decreased their alcohol consumption, strated that the intervention could but that patients who participated in At 1 year after the intervention,4 motivate alcohol-dependent patients the MI showed significantly greater participants in all three groups reported to initiate alcoholism treatment. These improvements in the following alcohol- having reduced their days of heavy investigators reported that 65 percent related variables: drinking, similar to the findings of the of patients with alcohol dependence study by Monti and colleagues (1999). who received the intervention and • They reported a significantly lower Moreover, the BIB group reported sig­ a direct referral from the ED to an incidence of drinking and driving nificantly fewer alcohol-related negative alcoholism treatment clinic kept their (62 percent, versus 85 percent in consequences (e.g., hangovers and lost initial appointment at the clinic, com­ the control group, which received work time) and alcohol-related injuries pared with 5.4 percent of the control standard care). than did the SC group. However, the group, who received just a referral. average number of injuries the partici­ Although a number of studies have • They were less likely to have been pants had sustained in the year preceding subsequently evaluated the effectiveness cited for a moving violation during the ED visit was low (i.e., an average of brief intervention in the ED, these the followup period (3 percent of 1.6 injuries), as was the incidence of are difficult to compare because they versus 23 percent). new injuries. Therefore, it is difficult to often used restricted populations (e.g., demonstrate significant changes, which only young adults), involved interventions • They were less likely to have sustained limits the interpretation of the findings. of varying lengths, and had method­ an alcohol-related injury during the Nevertheless, the investigators concluded ological limitations (D’Onofrio and followup period (21 percent versus that a booster added to a brief inter­ Degutis 2002). To date, only four studies 50 percent). vention with injured ED patients have analyzed the effectiveness of brief who engage in hazardous and harmful interventions in ED patients according • They reported fewer alcohol-related drinking may be helpful in reducing to stringent scientific standards (i.e., by social and legal problems in the 6 negative consequences and alcohol- including control groups and randomly months following treatment, such related injuries. assigning participants to an intervention as problems with dates, friends, or control group). All four studies, which parents, school, or the police. are described in the following sections, Study Conducted by Spirito and Colleagues included only injured patients. (For a Study Conducted by Longabaugh summary of the characteristics and and Colleagues Spirito and colleagues (2004) studied findings of these studies, see the table.) adolescents ages 13 to 17 who were Longabaugh and colleagues (2001) treated in an ED for an alcohol-related evaluated the effects of a brief motiva­ Study Conducted by Monti event. The adolescents were eligible to tional intervention in injured drinkers and Colleagues participate in the study if they had evi­ age 18 or older who visited the ED dence of alcohol in their blood, breath, Monti and colleagues (1999) compared of an urban teaching hospital. Patients the effectiveness of “standard care” with were eligible for the study if they screened 3The Alcohol Use Disorders Identification Test (AUDIT) is that of a brief motivational interview in positive for hazardous or harmful drink­ a commonly used screening tool to detect hazardous or reducing alcohol-related consequences ing (i.e., had breath alcohol concentra­ harmful drinking. A score of 8 points or more generally indicates that a patient is likely to meet the criteria for and alcohol use among ED patients tions greater than 0.003 mg/dL, reported hazardous or harmful drinking. ages 18 and 19. Standard care was having ingested alcohol in the 6 hours described as “consistent with general before the injury, or scored positive on 4For the followup, it was possible to contact 84 percent of 3 the original participants—39 percent in person and 61 practice for treating alcohol-involved the AUDIT screening test ). percent by mail or phone.

Vol. 28, No. 2, 2004/2005 65 or saliva (N = 142), or if they reported more, at the beginning of the study the Researchers interviewed the adoles­ drinking alcohol in the 6 hours before investigators administered the Adolescent cents by phone after 3 months and the injury that required treatment in Drinking Inventory (ADI) to identify contacted them in person after 6 and the ED (N = 10). The participants adolescents with potential alcohol prob­ 12 months. The investigators found underwent a battery of assessments that lems warranting a treatment referral and that adolescents in both groups drank took an average of 45 minutes to com­ for use in the personal feedback compo­ less alcohol during the 12-month fol­ plete. They reported their drinking nent of the intervention condition. The lowup period. However, adolescents behavior over the past 12 months and ADI is a 24-item measure of severity of in the MI group with a baseline ADI completed the Adolescent Drinking alcohol involvement, with a score > 15 score indicating problematic alcohol use Questionnaire (which assesses behavior indicating that referral for alcohol prob­ improved significantly in two outcomes, over the past 3 months), the Young Adult lems is needed. Participants then were average number of drinking days per Drinking and Driving Questionnaire, and randomly assigned to receive standard month (frequency) and frequency of the Adolescent Injury Checklist. Further­ care or a brief motivational intervention. high-volume drinking (bingeing). Based

Comparison of Four Clinical Studies Evaluating the Effectiveness of Brief Interventions in Emergency Departments and Inpatient Trauma Units*

Study Design Patient Population and Followup Study and Setting Admission Criteria Intervention Rate Outcome Effect

Monti et Design: RCT 94 patients ages 18–19, • Standard care • 3 months • Decrease in Positive al. 1999 Setting: ED treated at an ED after • One 35- to 40­ (phone): 93% alcohol con­ effect with an alcohol-related event minute BI (moti­ • 6 months (in sumption in the BI • Positive BAC or vational interview) person): 89% both groups • Report of drinking Interventions per­ • Greater reduc­ prior to the event that formed by 12 tion in alcohol- precipitated treatment experienced related injuries research assistants during the fol­ (bachelor’s and lowup period in master’s level) the BI group • Greater reduc­ No followup tion in other sessions alcohol-related problems (e.g., drinking and driving, social and legal prob­ lems) in the BI group

Gentilello Design: RCT 762 patients ages ≤ 18 • Standard care • 6 months: • Greater reduc- Positive et al. 1999 Setting: admitted to a trauma • One 30-minute BI 75% tion in alcohol- effect with Inpatient center (motivational • 12 months: related injuries the BI Trauma • BAC ≤ 100 mg/dL or interview) 54% during the fol­ Center • SMAST score ≤ 3 or Interventions per­ lowup period in • BAC 1–99 mg/dL and formed by one the BI group SMAST score of 1 or 2 Ph.D.-level psy­ • Greater decrease or chologist in alcohol con­ • BAC 1–99 mg/dL and Followup letter sumption in the elevated GGT or sent after 1 month BI group • SMAST score of 1 or • Greater reduc­ 2 and elevated GGT tion in ED visits and hospitaliza­ tions in the BI group

* RCT = randomized controlled trial, ED = emergency department, BAC = blood alcohol concentration, BI = brief intervention, BIB = brief intervention with booster, SMAST = Short Michigan Alcohol Screening Test, AUDIT = Alcohol Use Disorders Identification Test, GGT = gamma glutamyltransferase.

66 Alcohol Research & Health Screening and BI in the Emergency Department

on these findings, the investigators rec­ high refusal rates—that is, as many as about their alcohol consumption for ommend that adolescents who are 47 percent of patients in the studies fear of getting in trouble because of treated in the ED for an alcohol-related refused to participate. Refusal rates of underage drinking. injury should be screened for preexisting this magnitude can introduce signifi­ Second, the SC conditions in the stud­ alcohol problems and should receive a cant bias (e.g., only patients who have ies did not truly represent the standard brief intervention if the screen is positive. less severe problems or are more willing of care. Other studies have demonstrated to change their drinking behavior may that emergency practitioners rarely Limitations of These Studies agree to participate). Refusal rates in screen their patients for alcohol problems studies involving adolescents may be or provide any intervention (D’Onofrio Several methodological issues may have particularly high because parents may and Degutis 2002). In the three studies, influenced the results and limited the need to give consent for their children however, patients in the SC groups generalizability of the three studies to participate, and adolescents may not received at least brief advice and a described so far. First, all three had want their parents (or others) to know handout on avoiding drinking and

Comparison of Four Clinical Studies Evaluating the Effectiveness of Brief Interventions in Emergency Departments and Inpatient Trauma Units*

Study Design Patient Population and Followup Study and Setting Admission Criteria Intervention Rate Outcome Effect Longabaugh Design: RCT 539 patients ages ≤ 18 • Standard care 1 year (phone, • Greater reduc­ Positive et al. 2001 Setting: ED with evidence of harmful • One 40- to 60­ mail, in per­ tion in alcohol- effect with or hazardous drinking, minute BI son): 83% related injuries the BIB whose injury did not • One 40- to 60­ during the fol­ require hospitalization minute BI fol­ lowup period in • BAC ≤ 0.003 mg/dL or lowed by sched­ the BIB group • Report of alcohol use uled return visit • Decreases in 6 hours prior to injury (booster) 7–10 alcohol con­ or days later (BIB) sumption in all groups • AUDIT score ≤ 8 Interventions per­ • Greater reduc­ formed by 8 clini­ tion in alcohol- cally experienced related negative research assistants consequences (Ph.D., master’s, or in the BIB group bachelor’s level)

Spirito et Design: RCT Adolescents treated in • Standard care (5 • 3 months Greater reduction Positive al. 2004 Setting: ED in an ED after an alcohol- minutes) (phone): in frequency of effect with an urban level related event • One 35- to 45­ 93.4% drinking and the BI for 1 trauma • Positive for alcohol in minute BI (moti­ • 6 months binge drinking for problem center breath, saliva, or blood vational interview) (in person): patients with pre­ drinkers 89.5% existing problem­ or Interventions per­ • 12 months atic alcohol use • Self-reported alcohol formed by 12 clini­ (in person): in the BI group use 6 hours prior to cally experienced 89.5% injury research assistants Note: 47% of adoles­ (bachelor’s and cents asked to partici­ master’s level) pate refused No followup sessions

* RCT = randomized controlled trial, ED = emergency department, BAC = blood alcohol concentration, BI = brief intervention, BIB = brief intervention with booster, SMAST = Short Michigan Alcohol Screening Test, AUDIT = Alcohol Use Disorders Identification Test

Vol. 28, No. 2, 2004/2005 67 driving—minimal interventions that of the other three studies. However, were not followed up also decreased nonetheless go beyond the standard of this study included only hospitalized their alcohol consumption. care commonly seen in ED settings. trauma patients who were found to The difference between this and the Third, the screening and assessment exhibit unhealthy alcohol use based other three studies, which may improve of the participants may have acted like on screening and/or testing.5 It did not the generalizability of these results, is an intervention, as indicated by the fact include patients who were only treated the inclusion of a more credible control that participants in all study groups in the ED and released. The participating group. The control group in this study decreased their alcohol use. Assessment received minimal screening and assess­ questionnaires that take 30–45 minutes ment, which were less likely to have to complete may have an impact similar acted as an intervention and to have to that of brief interventions of similar Providing some confounded the results. Also, in contrast duration. Furthermore, questionnaires to the other studies, this investigation such as the Adolescent Injury Checklist, form of brief included patients who had sustained Adolescent Drinking Inventory, and injuries significant enough to warrant Drinking Inventory of Consequences intervention admission to the hospital, which in emphasize the consequences of alcohol to ED patients itself may lead to a so-called teachable misuse and in themselves may provide moment and may contribute to the feedback that motivates people to think whose injuries are patients’ motivation to change their about their behavior; such feedback is alcohol related may behavior. Limitations of the study one of the key components of the included a relatively high refusal rate intervention being tested. decrease their (34 percent of eligible patients did Fourth, to detect statistically signifi­ not participate) and a relatively low cant differences between control and alcohol consumption. followup rate. Moreover, patients who intervention groups, adequate numbers are hospitalized because of injuries of patients must participate and be make up only a small proportion of randomly distributed to the different patients represented the full spectrum the patients with alcohol problems groups. The three studies discussed, of unhealthy alcohol use. The patients who present to the ED for treatment of however, did not report conducting then were randomly assigned to either injuries, and the findings may therefore an analysis to determine if their sample the control group or the intervention not apply to all patients with alcohol size was sufficient to detect differences group, which received a single 30­ problems. (i.e., a power analysis). This makes minute motivational interview conducted interpretation of the studies’ findings by a doctoral-level psychologist in the Applicability of Study Findings difficult, particularly because none of inpatient setting. in Everyday ED Practice the studies found a significant differ­ Followup with the participants was ence between the intervention and conducted after 6 and 12 months. All four studies have suggested that control groups with respect to alcohol Although only 54 percent of participants providing some form of brief interven­ consumption. Without a power analy­ were available for the followup at 12 tion to ED patients whose injuries are sis, it is difficult to determine whether months, the investigators found that alcohol related may decrease their alcohol this lack of differences is genuine or those in the intervention group decreased consumption and alcohol-related nega­ just results from inadequate sample their weekly alcohol consumption sig­ tive consequences. The specific message sizes. Power analyses also should be nificantly more (by 21.8 drinks) than that should be delivered to the patients, conducted for analyses of the study the control group (by 6.7 drinks). The however, is not so clear, because the data on negative consequences of alco­ decrease was greatest in patients with standard care groups—which received hol consumption (e.g., drinking and mild to moderate alcohol problems at some brief advice, information, or driving or number of alcohol-related the beginning of the study. Furthermore, assessments containing motivational injuries). Such power analyses may the beneficial effects of the intervention statements—also experienced positive be challenging, however, because the appeared to be persistent, because after outcomes. Also, because the interventions prevalences of such consequences may a 3-year followup period, the investiga­ in these studies were implemented by be low at baseline. tors found a 47-percent reduction in research staff, who are not typically injuries requiring an ED visit or readmis­ available in most EDs (i.e., social workers, graduate students, or doctoral-level psy­ Study Conducted by Gentilello sion to the trauma service in the inter­ vention group (Gentilello et al. 1999). chologists), it is unclear how the findings and Colleagues Because of the low followup rate, how­ can be translated into the real-world The fourth study, conducted by Gentilello ever, it is not possible to generalize the 5Screening involved completion of a questionnaire, and colleagues (1999), avoided at least findings of this study, as one cannot whereas testing refers to analyses of alcohol levels in the some of the methodological limitations determine whether the patients who blood, saliva, or breath.

68 Screening and BI in the Emergency Department

ED setting. Other research on the fea­ residents contained evidence of screen­ brief intervention for alcohol problems sibility of screening and brief interven­ ing and intervention, compared with in the ED as well as in other health care tions in the ED setting can shed addi­ 17 percent of records of patients settings. One of the most important tional light on this question. treated by a control group of similar issues is the relationship between patient Degutis (1998) demonstrated that residents who did not receive training. and practitioner, and the practitioner’s screening with tools such as quantity/ responsibility to identify behaviors such frequency questions and the four-item as unhealthy alcohol use that can result CAGE questionnaire was feasible in a Challenges Associated in health problems. Identifying such real-world ED setting. (For information With Screening and behaviors may create a dilemma for on the CAGE and other commonly used Intervention in the practitioners who are uncertain about screening instruments, see the sidebar ED Setting how to intervene and where to refer “Screening Tests,” on page 78 of this these patients without compromising issue.) Similarly, Hungerford and col­ Many barriers to screening, brief inter­ the patient–physician relationship. leagues (2003), using a study popula­ vention, and referral have been identified Nevertheless, practitioners have an tion of young adults ages 18 to 39, in the ED setting. This environment is obligation to identify health risks asso­ reported that screening and interven­ always chaotic, and time is precious. ciated with alcohol use and to treat tion could be integrated in the ED Lack of confidence on the part of the the alcohol use problems themselves or setting. In this study, research staff emergency practitioners regarding their refer the patients to the appropriate screened a convenience sample6 of ED ability to screen patients effectively, resources, just as they would do for any patients who were waiting for treat­ scarcity of role models who are perform­ other chronic disease, such as hyperten­ ment. The investigators found that 87 ing screening, and inadequate resources sion or diabetes (McLellan et al. 2000). percent of the young adult drinkers often are cited as reasons why practi­ To illustrate, screening for alcohol prob­ consented to screening. Of these, 43 tioners fail to screen and intervene. In lems often is compared with screening percent screened positive for alcohol addition, ethical matters and insurance for tetanus immunization. Every injured problems on the AUDIT,7 and of those constraints may present obstacles. ED patient is asked whether he or she with positive screens, 94 percent received has been vaccinated against tetanus and counseling. The high prevalence of Time Constraints is offered treatment, no matter how alcohol problems and the broad accep­ busy the practitioner is. Even though tance of screening and brief interven­ To identify potential barriers more most practitioners have never even seen tion in this sample indicated that even accurately, Graham and colleagues (2000) a case of tetanus, they allocate time for though the study used research staff, surveyed 569 members of the Michigan this screening. Therefore, considering who are not present in real EDs, College of Emergency Physicians about the high number of ED patients suffer­ screening is feasible in this setting, their attitudes toward using interventions ing from alcohol-related injuries and and the ED is a promising venue for with ED patients who have alcohol other adverse consequences, it is diffi­ screening and brief intervention. problems. Of the 257 members who cult to understand why practitioners A survey of emergency practitioners responded (46 percent of those sur­ often ignore this health problem, (D’Onofrio et al. 2004) found that veyed), 75 percent agreed that alcohol which they typically encounter during these clinicians considered performing abuse and dependence are treatable ill­ each shift. a brief intervention for harmful and nesses, and only 15 percent stated they In order to make the process of hazardous drinkers feasible and accept­ would not support ED interventions. screening, intervention, and referral as able in their everyday practice. Other Both supporters and nonsupporters easy as possible and thereby promote its investigators demonstrated that emer­ thought that lack of time was a major use, the American College of Emergency gency medicine residents who received obstacle to screening and intervention. Physicians (2004) has developed an training in screening and brief inter­ Consequently, the study’s authors sug­ Alcohol Screening and Brief Intervention vention in a skills-based workshop gested that existing interventions be Resource Kit that is available via the increased their knowledge and practice adapted to the time limitations of the Internet. This toolbox contains an of these procedures (D’Onofrio et al. ED setting. One type of brief interven­ explanation of brief interventions, samples 2002). Fifty-eight percent of medical tion that can be performed in less than of patient handouts, and information records of patients treated by trained 10 minutes already has been developed about how to develop resource lists for and tested specifically for emergency individual communities. 6The term “convenience sample” means that the investi­ practitioners (D’Onofrio et al. 1998). gators picked potential participants at will, rather than just using consecutive patients or another preset selection Insurance Coverage scheme. This process may introduce some bias into the Ethical and Legal Issues sample. Some health practitioners also are reluc­ 7In this study, a score of > 6 points was considered a Several ethical and legal issues frequently tant to screen their patients for alcohol positive screen. arise in the discussion of screening and use because they are concerned that if

Vol. 28, No. 2, 2004/2005 69 they identify an alcohol problem, the diseases—still remains. A Federal law of having two or more drinks (11 per­ patients’ health insurance carrier may requiring insurance coverage parity for cent). In addition, 95 percent of patients deny reimbursement for the ED visit. AOD abuse treatment was first intro­ in the intervention group requested This concern is especially prevalent in duced in both Houses of Congress in further health information. One week ED and trauma care settings, because 1997 but has yet to pass. A few States after the ED visit, 62 percent of the many States have insurance regulations have laws requiring parity, but these laws intervention group remembered receiv­ allowing insurers to exclude coverage ing advice on what they could do to for a loss sustained because the insured improve their health. The investigators was under the influence of alcohol and concluded that computer technology other drugs (Rivara et al. 2000). These may help physicians use the patients’ exclusions are based on a widely adopted Computer technology waiting time for health promotion and model called the Uniform Accident may help physicians to target patients at risk for various and Sickness Policy Provision Law. (For health problems. more information on the denial of use the patients’ Gregor and colleagues (2003) exam­ health care coverage for ED visits for waiting time for ined the feasibility of using an interac­ AOD-related problems, see the accom­ tive computer program in the ED to panying sidebar by Chezem.) Only two health promotion prevent alcohol misuse among adoles­ States (Iowa and South Dakota) specif­ cents, enrolling patients ages 14 to 18 ically prohibit insurers from denying and to target patients who visited the ED within 24 hours of coverage in this situation. at risk for various an acute injury. Of the participants, 71 The arguments related to denial of percent reported ever drinking alcohol, insurance coverage, however, do not seem health problems. and about 63 percent reported recent justifiable in the context of screening alcohol use. The program consisted of because screening differs from testing an interactive house party with audio. (e.g., determination of blood alcohol Each participant chose a “party pal” levels). To deny reimbursement, the only apply to insurance plans that are from a group of five teenaged cartoon insurer must demonstrate that AOD use regulated by the States and do not include characters and was exposed to various to some degree caused the insured’s injury Federally regulated plans (e.g., the Federal scenarios depicting important concepts or other problem. This demonstration Employees Health Benefit Plan) that regarding alcohol misuse. Of the recent generally is based on a diagnostic or are governed by Employment Retirement drinkers participating in the study, 74 laboratory test documenting a specific Income Security Act (ERISA) statutes. percent reported that the program made AOD concentration in the patient’s them rethink their alcohol use, 94 per­ blood, which is linked to the reason for cent liked the program, and only about the ED visit or hospitalization. Screening Innovative Approaches to 5 percent required assistance with it. using structured questionnaires, in con­ Screening in the ED An ongoing study currently is evaluat­ trast, identifies an existing problem with­ ing the effectiveness of the program in out linking it directly to the patient’s Several studies have reported innovative reducing participants’ alcohol-related current visit and therefore provides an methods for screening and intervention problem behaviors. opportunity to intervene. Thus, it prob­ in the ED. Rhodes and colleagues (2001) Another approach, being studied at ably would be difficult for an insurer to described the use of a computer-based Boston Medical Center and Yale–New connect the fact that a person drinks approach to screening and general health Haven Hospital, is the Project ASSERT more than recommended by national promotion in the ED. (This approach model, which uses Health Promotion drinking guidelines to an ED visit for was not specific to AOD-related prob­ Advocates (HPAs) or community out­ a wrist fracture that occurred when the lems but addressed a variety of health reach workers to screen, intervene, and patient slipped on ice going to work. issues.) In this study, 542 adult ED refer patients with alcohol problems. Consistent with this assumption, all patients with nonurgent conditions Bernstein and colleagues (1996) evalu­ legal cases related to the denial of cov­ (89 percent of those approached) were ated the effectiveness of this program erage have been based on testing that assigned either to the computer inter­ in 245 participants, most of whom was done at the time of an incident but vention—a self-administered computer had alcohol dependence. After 90 days, were not based on screening for a pre­ survey generating individualized health these participants reported a 56-percent existing alcohol problem. information—or to usual care (i.e., no reduction in alcohol use and a 64­ The larger issue of substance abuse intervention). In the intervention group, percent reduction in binge drinking. treatment parity—that is, regulations 85 percent of participants reported one In addition, more than 50 percent of requiring that alcohol and other drug or more behavioral risk factors for AOD participants had received a treatment abuse treatment be reimbursed at problems, including problem drinking referral. In another evaluation of Project the same level as treatment for other (19 percent), or driving within 4 hours ASSERT, 10,572 patients were screened

70 Alcohol Research & Health Screening and BI in the Emergency Department

and evaluated, and 1,343 patients were by phone or in person, or whether makers still must address a variety of referred to specialized AOD treatment self-reported outcomes should be research questions and policy issues. ■ programs over a 2-year period (D’Onofrio obtained by phone, in person, or by and Degutis 2003). HPAs were subse­ interactive voice response. quently able to contact 811 of the References referred patients, of which 711 (88 •Explore how screening and interven­ American Psychiatric Association (APA). Diagnostic percent) had enrolled in a treatment tion can best be incorporated into and Statistical Manual of Mental Disorders, Fourth program. The results suggest that this clinical practice, whether existing Edition. Washington, DC: APA, 1994. model of screening and direct linkage ED staff (e.g., nurses, physicians, American College of Emergency Physicians. Alcohol to treatment is feasible. Moreover, physician extenders) can be used at Screening and Brief Intervention Resource Kit. ACEP the program is likely to be sustainable, least to some extent, and what other Product No. 409036. Available at: http://www. because Project ASSERT is funded by alternatives (e.g., peer educators, acep.org (follow link to Practice Resources). HPA consultation fees that are included interactive computerized methods, Accessed Sept. 26, 2004. in the hospitals’ billing processes. or volunteers) are available. BERNSTEIN, E.; BERNSTEIN, J.; AND LEVENSON, S. Project ASSERT: An ED-based intervention to • Establish the cost–benefit ratios of increase access to primary care, preventive services, and the substance abuse treatment system. Annals Future Directions various interventions, determine the of Emergency Medicine 30:181–189, 1996. most effective method of recovering The studies discussed in this article costs, and explore how practitioners Centers for Disease Control and Prevention (CDC), National Center for Injury Prevention and Control. have suggested that screening and brief can be adequately reimbursed for 10 Leading Causes of Death. Available online at intervention in the ED can be feasible any counseling they provide. http://www.cdc.gov/ncipc/osp/charts.htm. Accessed and effective, but more research is nec­ Sept. 26, 2004. essary to determine optimal treatment •Investigate how to overcome obstacles CHAFETZ, M.E.; BLANE, H.T.; ABRAMS, H.S.; ET approaches and their implementation to screening and intervention, such AL. Establishing treatment relations with alcoholics. in the ED setting. Such research should as difficulties with insurance cover­ Journal of Nervous and Mental Disease 134:395– age and unwillingness or inability 409, 1962. •Include both injured and noninjured of practitioners to perform these CHERPITEL, C.J. Drinking patterns and problems: patients to reflect the heterogeneous measures. A comparison of primary care with the emergency patient population seen in EDs. room. Substance Abuse 20:85–95, 1999. •Address how screening and brief DEGUTIS, L.C. Screening for alcohol problems in •Include sample sizes adequate to detect intervention skills can be incorporated emergency department patients with minor injury: differences between intervention and in the practitioners’ initial profes­ Results and recommendations for practice and pol­ control groups on key outcome vari­ sional education. icy. Contemporary Drug Problems 25:463–475, 1998. ables such as alcohol consumption and D’ONOFRIO, G., AND DEGUTIS, L.C. Preventive other negative consequences. • Assess screening and intervention care in the emergency department: Screening and for culturally diverse groups and brief intervention for alcohol problems in the emer­ gency department: A systematic review. Academic •Include control groups exposed to non-English-speaking patients, who Emergency Medicine 9:627–638, 2002. minimal assessments so as to mini­ represent a significant proportion of mize the potential influence of the ED patients but have not been a D’ONOFRIO, G., AND DEGUTIS, L.C. Linking emergency department patients with alcohol and assessment on the patients and thus focus of existing studies. other drug problems to treatment (Abstract). compare interventions with true Academic Emergency Medicine 10:515, 2003. “usual care.” D’ONOFRIO, G.; BERNSTEIN, E.; BERNSTEIN, J.; ET AL. Patients with alcohol problems in the emer­ •Evaluate specific components of Conclusions gency department. Part 2: Intervention and referral. interventions (e.g., motivational Academic Emergency Medicine 5:1210–1217, 1998.

enhancement or simple feedback ED visits provide health care practitioners D’ONOFRIO, G.; NADEL, E.S.; DEGUTIS, L.C.; ET AL. and advice) to determine which with an important opportunity to screen Improving emergency medicine residents’ approach components are most effective in their patients for alcohol problems and, to patients with alcohol problems: A controlled edu­ the ED setting. Such analyses could if necessary, to initiate brief interven­ cational trial. Annals of Emergency Medicine 40:50–62, clarify whether intervention effec­ tion. Research has demonstrated that 2002. tiveness depends on the patients’ screening and brief interventions are D’ONOFRIO, G.; PANTALON, M.; DEGUTIS, L.C.; readiness to change and/or the feasible and effective in the ED setting. ET AL. Teaching brief intervention for alcohol prob­ lems to emergency practitioners: Development, severity of their injury or illness, However, to be successful in changing implementation, and testing of a model. Annals of whether “boosters” are beneficial physicians’ daily practices and decreasing Emergency Medicine 44(Suppl.):S78, 2004. and cost-effective, whether followup the harmful consequences of alcohol GENTILELLO, L.M.; RIVARA, F.P.; DONOVAN, D.M.; and boosters should be conducted misuse, clinicians, researchers, and policy- ET AL. Alcohol interventions in a trauma center as a

Vol. 28, No. 2, 2004/2005 71 means of reducing the risk of injury recurrence. patients to primary care. Academic Emergency Medicine RIVARA, F.P.; TOLLEFSON, S.; TESH, E.; AND Annals of Surgery 230:473–484, 1999. 9:639–642, 2002. GENTILELLO, L.M. Screening trauma patients for alcohol problems: Are insurance companies barri­ GRAHAM, D.M.; MAIO, R.F.; BLOW, F.C.; AND MCLELLAN, A.T.; LEWIS, D.C.; O’BRIEN, C.P.; ers? Journal of Trauma: Injury, Infection, and HILL, E.M. Emergency physician attitudes concern­ AND KLEBER, H.D. Drug dependence, a chronic Critical Care 48:115–118, 2000. ing intervention for alcohol abuse/dependence deliv­ medical illness: Implications for treatment, insur­ ered in the emergency department: A brief report. ance, and outcomes evaluation. JAMA: Journal of the ROCKETT, I.R.; PUTNAM S.; JIA, H.; AND SMITH, G. Assessing substance abuse treatment need: A Journal of Addictive Diseases 19:45–53, 2000. American Medical Association 284:1689–1695, 2000. statewide hospital emergency department study. GREGOR, M.A.; SHOPE, J.; BLOW, F.; ET AL. Feasibility MILLER, W.R., AND ROLLNICK, S. Motivational Annals of Emergency Medicine 41:802–813, 2003. of using an interactive laptop program in the emer­ Interviewing: Preparing People to Change Addictive gency department to prevent alcohol misuse among SAITZ, R. Clinical Practice. Unhealthy alcohol use. adolescents. Annals of Emergency Medicine 42:276– Behavior. New York: Guilford, 1991. New England Journal of Medicine 352(6):596–607, 2005. 284, 2003. MONTI, P.M.; COLBY, S.M.; BARNETT, N.P.; ET AL. HUNGERFORD, D.; WILLIAMS, J.; FURBEE, P.; ET AL. Brief intervention for harm reduction with alcohol- SPIRITO, A.; MONTI, P.M.; BARNETT, N.P.; ET AL. Feasibility of screening and intervention for alcohol positive older adolescents in a hospital emergency A randomized clinical trial of a brief motivational problems among young adults in the ED. American department. Journal of Consulting and Clinical intervention for alcohol-positive adolescents treated Journal of Emergency Medicine 21:14–22, 2003. Psychology 67:989–994, 1999. in an emergency department. Journal of Pediatrics 145:396–402, 2004. Institute of Medicine. Broadening the Base of National Highway Transportation Safety Treatment for Alcohol Problems: Report of a Study by Administration (NHTSA), National Center for Substance Abuse and Mental Health Services a Committee of the Institute of Medicine, Division of Statistics and Analysis. Alcohol Involvement in Fatal Administration (SAMHSA). Overview of Findings Mental Health and Behavioral Medicine. Washington, Motor Vehicle Traffic Crashes, 2003. Pub. No. From the 2002 National Survey on Drug Use and Health. NHSDA Series H–21, DHHS Pub. No. DC: National Academy Press, 1990. DOT HS–809–822. Washington, DC: U.S. Dept. SMA 03–3774. Rockville, MD: Office of Applied of Transportation, NHTSA, 2005. LONGABAUGH, R.; WOOLARD, R.F.; NIRENBERG, Studies, 2003a. ET AL T.D.; . Evaluating the effects of a brief moti­ RHODES, K.V.; LAUDERDALE, D.S.; STOCKING, vational intervention for injured drinkers in the Substance Abuse and Mental Health Services C.B.; ET AL. Better health while you wait: A con­ emergency department. Journal of Studies on Administration (SAMHSA). The NHSDA Report: trolled trial of a computer-based intervention for Alcohol 62:806–816, 2001. Alcohol Use by Persons Under the Legal Drinking Age screening and health promotion in the emergency of 21. NHSDA Series H–13, DHHS Pub. No. MCCARTHY, M.L.; HIRSHON, J.M.; RUGGLES, department. Annals of Emergency Medicine 37:284– SMA 01–3549. Rockville, MD: Office of Applied R.L.; ET AL. Referral of medically uninsured ED 291, 2001. Studies, 2003b.

72 Alcohol Research & Health