Stout, R.L. Prevention experiments in the context of ongoing community process: Opportunities or obstacles for research. In: Holder, H.D., and Howard, J., eds. Methodological Issues in Community Prevention Trials for Alcohol Problems. Westport, Conn.: Praeger Publishing, in press.

Strickland, D.E. Advertising exposure, alcohol consumption, and misuse of alcohol. In: Grant, M.; Plant, M.; and Williams, A., eds. Economics and Alcohol: Consumption and Controls. New York: Gardner Press, 1983. pp. 201–222. Sue, S.; Zane, N.; and Ito, J. Alcohol drinking patterns among Asian and Caucasian Americans. J Cross-cultural Psychol 10(1):41–56, 1979. Tobler, N.S. Meta-analysis of 143 adolescent drug prevention programs: Quantitative outcome results of program participants compared to a control or comparison group. J Drug Issues 16:537-567, 1986.

Treno, A.J.; Parker, R.N.; and Holder, H.D. Understanding U.S. alcohol consumption with social and economic factors: A multivariate time series analysis, 1950-1986. J Stud Alcohol, in press. Trice, H.M., and Beyer, J.M. Work-related outcomes of the constructive-confrontation strategy in a job-based alcoholism program. J Stud Alcohol 45:393–404, 1984.

Trice, H.M., and Sonnenstuhl, W.J. Drinking behavior and risk factors related to the work place: Implications for research and prevention. J App Behav Sci 24(4):327–346, 1988.

U.S. General Accounting Office. Drinking-Age Laws: An Evaluation Synthesis of Their Impact on Highway Safety. Washington, D.C.: U.S. General Accounting Office, 1987.

U.S. Indian Health Service. Chart Book Series. Washington, D.C.: U.S. Department of Health and Human Services, 1988.

Van Dyke, J., and Womble, K. The Impact of Minimum Drinking Age Laws on Fatal Crash Involvements: An Update of the NHTSA Analyses. Washington, D.C.: National Highway Traffic Safety Administration, 1988. Wagenaar, A.C., and Holder, H.D. Effects of alco

holic beverage server liability on traffic crash injuries. Alcohol Clin Exp Res 15(6):942–947, 1991. Wallack, L. Alcohol advertising reassessed: The public health perspective. In: Grant, M.; Plant, M.; and Williams, A., eds. Economics and Alcohol: Consumption and Controls. New York: Gardner Press, 1983. pp. 243–248. Waterson, E.J., and Murray-Lyon, I.M. Preventing alcohol-related birth damage: A review. Soc Sci Med 30(3):349–364, 1990.

Wechsler, H.; Kasey, E.H.; Thum, D.; and Demone,
H.W., Jr. Alcohol level and home accidents. Pub-
lic Health Rep 84(12):1043-1050, 1969.
Wechsler, H., and Thum, D. Teenage drinking,

drug use, and social correlates. J Stud Alcohol
34:220-227, 1973.

Welte, J.W., and Abel, E.L. Homicide: Drinking by
the victim. J Stud Alcohol 50(3):197–201, 1989.
Welte, J.W.; Abel, E.L.; and Wieczorek, W. The role
of alcohol in suicides in Erie County, NY, 1972–
84. Public Health Rep 103(6):648-652, 1988.
White, H.R.; Pandina, R.J.; and LaGrange, R.L. Lon-
gitudinal predictors of serious substance use
and delinquency. Criminology 25(3):715-740,
1987.

Williams, T.P., and Lillis, R.P. Long-term changes in reported alcohol purchasing and consumption following an increase in New York State's purchase age to 19. Br J Addict 83(2):209–217, 1988. Wilsnack, S.C. Drinking, sexuality, and sexual dys

function in women. In: Wilsnack, S.C., and Beckman, L.J., eds. Alcohol Problems in Women: Antecedents, Consequences, and Interventions. New York: Guilford Press, 1984. Womble, M. Black women. In: Engs, R.C., ed.

Women: Alcohol and Other Drugs. Dubuque, Iowa: Kendall/Hunt Publishing Company, 1990. pp. 127-135.

Worden, J.K.; Flynn, B.S.; Merrill, D.G.; Waller, J.A.; and Haugh, L.D. Preventing alcohol-impaired driving through community self-regulation training. Am J Public Health 79(3):287–290, 1989. Wright, R., Jr., and Watts, T.D., eds. Prevention of Black Alcoholism: Issues and Strategies. Springfield, Ill.: Charles C. Thomas, 1985. Wylie, A.; Casswell, S.; and Stewart, J. The response of New Zealand boys to corporate and sponsorship alcohol advertising on television. Br J Addict 84(6):639–646, 1989.

Young, S.L. Increasing the noticeability of warn

ings: Effects of pictorial, color, signal icon and border. Proceedings of the Human Factors Society 35th Annual Meeting, San Francisco, Calif., September 1991, pp. 580–584.

Zador, P.L.; Lund, A.K.; Fields, M.; and Weinberg,

K. Fatal Crash Involvement and Laws Against Alcohol-impaired Driving. Washington, D.C.: Insurance Institute for Highway Safety, 1988. Zambrana, R.E., and Aguirre-Molina, M. Alcohol abuse prevention among Latino adolescents: A strategy for intervention. J Youth Adolescence 16(2):97-113, 1987.

Zuckerman, M. Sensation Seeking: Beyond the Optimal Level of Arousal. Hillsdale, N.J.: Lawrence Erlbaum Associates, 1979.

Introduction

T

his chapter focuses on screening and brief intervention procedures used to detect alcohol problems in individuals. Screening involves the use of easily administered, inexpensive procedures to identify people at risk for alcoholism and those who may be developing alcohol problems. The goal of screening is early identification and thus prevention through early intervention.

Brief intervention procedures are time-limited therapeutic strategies. For alcohol use problems, such interventions generally range from one to five sessions, each lasting no longer than an hour (Babor and Grant 1989). The goal of brief intervention strategies is prevention, that is, changing the behavior of individuals who are experiencing adverse effects of drinking but who are not physically dependent on alcohol. The brief intervention procedures discussed in this chapter include those tested and applied in health care settings and programs conducted under the auspices of courts, schools, and workplaces. Most of the research studies cited have been conducted in health care settings. These settings serve as a unique environment that provides an opportunity to identify many persons adversely affected by alcohol use and to implement therapeutic interventions.

Screening

Screening activities that attempt to identify individuals who may have or may develop an alco

hol problem differ from assessment (Allen et al. 1988). Whereas the goal of screening is to detect potential or developing alcohol problems, the goals of assessment are to establish a diagnosis and aid in the establishment of a specific treatment plan. Screening procedures are typically brief and can be conducted in a variety of settings by persons with limited clinical experience. Assessments, on the other hand, are usually conducted by alcohol and drug abuse specialists over multiple visits in a treatment center. Figure 1 illustrates the role of screening in the prevention of alcohol problems.

The goal of brief intervention strategies is prevention, that is, changing the behavior of individuals who are experiencing adverse effects of drinking but who are not physically dependent on alcohol.

Screening programs can be divided into two major types. The first includes programs directed at asymptomatic persons such as those who seek medical care for problems unrelated to alcohol use, those who participate in random alcohol and/or drug screening programs, or drivers who are stopped at random for breathalyzer testing. The second type consists of programs directed toward identified individuals who might have alcohol-related problems. Examples include hospital-based programs that screen patients

admitted for traumatic injuries (Graham 1991), court-mandated programs to screen adolescents arrested for violent crimes, breathalyzer testing conducted by the police on persons suspected of driving while intoxicated, and employee assistance programs that primarily screen workers with impaired performance.

Screening programs of the second type are accepted and well established; screening asymptomatic persons is less well established, but data exist that suggest the potential of such efforts. For example, Umbricht-Schneiter et al. (1991) concluded that routine screening of hospitalized persons for alcohol problems could improve medical care after they screened admissions and found that the prevalence of alcohol abuse among hospitalized patients (22.4 percent) was much higher than reflected in discharge diagnoses (7.4 percent). There are no national efforts to identify asymptomatic individuals with alcohol problems that are comparable in scope to national screening programs for diabetes, hypertension, high cholesterol, and cancer. Some possible reasons for this situation include ethical dilemmas, confidentiality issues, costs, lack of scien

tific evidence to support mass screening, absence of foolproof methods, systems barriers, and societal drinking norms.

Scientific Rationale for Screening

The Canadian Task Force on Periodic Health Examination (1979), Frame (1986), and the U.S. Preventive Services Task Force (1989) examined the scientific evidence supporting routine screening for alcohol problems and used a number of criteria to make their recommendations: (1) Is the incidence high enough to justify the cost of screening? (2) Does the problem have a significant effect on the quality or quantity of life? (3) Is effective treatment available? (4) Are screening tests available that are valid and costeffective? (5) Are the adverse effects of screening tests acceptable to clinicians and patients? (6) Does early treatment in the asymptomatic period reduce alcohol-related morbidity and mortality?

The first four criteria have been established, as demonstrated by reviews in this Special Report to the U.S. Congress on Alcohol and Health and in the Institute of Medicine report (1990). Alco

hol problems are common; they are associated with serious health and social consequences; effective treatment is available; and there are valid, cost-effective procedures for screening.

The one criterion that has not been fully established to justify screening asymptomatic individuals is the effectiveness of screening in reducing alcohol-related morbidity and mortality. Although early identification and treatment of alcohol problems may reduce the number of highway fatalities, accidents, injuries, and employment-related problems (Kristenson et al. 1983), there have been no long-term studies showing that early treatment is significantly more effective in reducing mortality due to alcoholrelated medical conditions than waiting until symptoms of alcohol dependence develop. Longterm prospective treatment studies, such as those completed for hypertension, high cholesterol, and breast cancer, have not yet been conducted for alcohol dependence.

Assessing the Usefulness of
Screening Tests

The usefulness of any screening test depends on its validity and reliability. Validity is concerned with whether the test actually measures what it is supposed to measure; reliability pertains to consistency of test results across raters and time. Related to validity are two key properties of every screening test, sensitivity and specificity (Sackett 1992). Sensitivity indicates how well a test detects a target disorder when it is present; the closer to 100 percent of those with alcohol problems who test as positive, the more sensitive the test. Specificity refers to a test's effectiveness in identifying people who do not have the disorder; the closer to 100 percent of persons without alcohol problems who test as negative, the more specific the test (Sackett 1992).

It is not, however, possible to maximize both values; false positives increase as sensitivity rises; missed cases (i.e., false negatives) increase with specificity (Rice 1987). Thus knowledge of a test's sensitivity and specificity alone is not sufficient in determining its usefulness. If the target disorder occurs only infrequently in a population, most of the cases identified by a sensitive test will be false positives (Rice 1987); thus the prevalence of a particular condition in the screened population must be taken into account (Grant et al. 1989).

Illustrating the impact of prevalence on the validity of screening tests are validation studies of the short (13-item) version of the Michigan Alcoholism Screening Test (SMAST) (Selzer et al. 1975) conducted by Fleming and Barry (1989b, 1991). In two samples of alcohol-dependent persons in a treatment program, the sensitivity of the SMAST was .94 to .98 (Fleming and Barry 1989b). Later, the authors (1991) administered the instrument to a sample of persons attending ambulatory medical care clinics; subjects were assessed for alcohol abuse and alcohol dependence according to Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R) (American Psychiatric Association 1987) criteria; 14 percent met the criteria for both. In this community sample, the sensitivity was only .56 when the same weighted cutoff score was used to indicate a positive test. Fleming and Barry (1989b, 1991) demonstrated that improvements can be made in the sensitivity and specificity of screening tests by altering item weighting and cutoff scores.

The usefulness of any screening test depends on its validity and reliability. Validity is concerned with whether the test actually measures what it is supposed to measure; reliability pertains to consistency of test results across raters and time.

The development of a “gold standard” that can be used to determine the predictive validity of screening tests is complicated by the heterogeneity of alcohol disorders (Babor 1990; Kosten and Kosten 1990; Tarter 1992) and the absence of biological measures. Although several laboratory tests are available, the alcohol field has no equivalent to a blood sugar measurement to identify diabetes or a blood pressure measurement to recognize hypertension. Finding a biological measurement that will reliably identify alcohol dependence remains a priority for alcohol research.

Clinicians and researchers have assumed that direct questions about alcohol consumption may have limited clinical value because people may tend to deny having alcohol use problems. However, a number of reviews on the validity of selfreports conclude that they are an important source of data that provide unique information

(Babor et al. 1987, 1990; Maisto et al. 1990; Midanik 1988, 1989; Sobell and Sobell 1990). Validity is contingent on several factors, including recent alcohol consumption, associated medical and psychiatric disorders, level of cognitive function, rapport between the client and interviewer, client awareness of corroborative lab tests or family member reports, appropriate consumption questions, and assurance of client confidentiality (Skinner 1984).

Methods to minimize the problem of inaccurate reporting include asking about specific amounts of drinking rather than average amounts; defining a single drink; inquiring about specific amounts of beer, wine, and hard liquor, and inquiring about frequency, quantity, and occasions of heavier use with separate questions.

The most common alcohol screening questions used by health care professionals focus on the quantity and frequency of alcohol use. Although there has been limited research on the most effective quantity and frequency questions, a number of questions have been recommended (Babor et al. 1987; Brown 1992; Skinner 1990). Examples include "How many days per week do you drink?” “On a day when you drink alcohol, how many drinks do you have?" In addition to determining quantity and frequency, it is important to assess patterns of heavier drinking, since episodes of heavy use can be associated with motor vehicle accidents, trauma, fights, and other adverse consequences. Therefore, questions on occasions of heavy drinking are contained in a number of tests, for example, "How many times in the last month did you drink more than five drinks at one sitting?"

Efforts to validate screening tests have been made, and specific techniques for questioning people about the quantity and frequency of their drinking have been developed. A number of investigators have used a 7-day drinking history to determine the predictive validity of screening tests (Cutler et al. 1988; Russell et al. 1991; Sokol et al. 1989). The 7-day Time Line Follow Back procedure has been shown in research settings to be a valid indicator of alcohol use (Babor et al. 1987; Sobell and Sobell 1975; Waterton and Duffy 1984). This procedure allows an inter

viewer to inquire about the type of beverage and quantity of alcohol use during three periods of time (morning, afternoon, and evening) for each of the previous 7 days. Cyr and Wartman (1988) have shown that asking about both alcohol use during the previous 24 hours and history of problem use was as accurate as administering the full 24-question Michigan Alcoholism Screening Test (MAST) (Selzer 1971). Other researchers have found a high degree of agreement between self-reports of recent alcohol use and breathalyzer testing in emergency department patients (Cherpitel 1989; Fine et al. 1978; Gibb et al. 1984).

Studies have shown that the use of sound techniques can maximize the validity of selfreports (Babor et al. 1987; Room 1991; Skinner 1984). Methods to minimize the problem of inaccurate reporting include asking about specific amounts of drinking rather than average amounts; defining a single drink; inquiring about specific amounts of beer, wine, and hard liquor; and inquiring about frequency, quantity, and occasions of heavier use with separate questions (Cutler et al. 1988; Sobell and Sobell 1990).

Specific Screening Tests

This section discusses specific screening instruments used to detect alcohol problems. Table 1 summarizes this information and shows recommended cutoff scores, populations for whom the test is recommended, and general comments.

CAGE

The test most widely used as a standard screening test in clinical practice is the CAGE (Ewing 1984; Mayfield 1974) (see also Buchsbaum et al. 1991; Bush et al. 1987; King 1986). The test assesses four areas related to lifetime alcohol use: Have you ever felt the need to CUT down on your drinking? Have you ever felt ANNOYED by someone criticizing your drinking? Have you ever felt GUILTY about your drinking? Have you ever felt the need for an EYE opener? Two positive responses are considered a positive test and indicate that further assessment is warranted. Like most of the screening instruments discussed, the sensitivity and specificity of the CAGE varies from 60 to 95 percent and 40 to 95 percent, respectively (Beresford et al. 1990; Bush et al. 1987). Such variability may be related to different criteria, assessment of lifetime use as compared with current use, variations in the cut