Preventive Intervention Cost-Effectiveness and Cost Benefit
Literature Review
August, 1998

Lisa Werthamer, M.S.W., Sc.D.
The University of Tennessee, Knoxville

and

Pinka Chatterji, M.A.
Johns Hopkins University School of Public Health

Hyperlinks to Sections Within This Text:

Authors' Note
Executive Summary
Introduction and Background
Theoretical Framework
Application
Future Directions
Conclusions
End Notes
References
Acknowledgments

AUTHORS' NOTE

    We sought to retrieve and critically analyze literature about the extent and magnitude of drug use, evaluation studies of drug prevention effectiveness, and original studies and reviews concerning the costs and benefits and cost-effectiveness of drug abuse prevention initiatives, other health and social prevention initiatives, and drug abuse treatment. We identified and retrieved literature from the following online databases: Current Contents (Social and Behavioral Sciences), GPO Monthly Catalog, MEDLINE, PsychINFO, Public Affairs Information Services, Sociological Abstracts, Social Work Abstracts, and General Economic Indices. Database searches were supplemented by reviewing drug abuse newsletters and abstracting journals including Addiction Abstracts, Alcoholism and Drug Abuse Weekly, Brown University Digest of Addiction Theory and Application, Substance Abuse Letter, and Substance Abuse Report.
    We used the keywords tobacco, alcohol, drugs, substances, abuse, and addiction to identify articles about the extent and magnitude of drug abuse, common interventions to prevent or reduce drug abuse, and cost analyses of drug abuse interventions. We used the keywords outcomes, effectiveness, costs, and benefits to identify articles concerning economic evaluations of health policies and programs. Reference lists of retrieved articles were reviewed to identify additional relevant citations, and all references were entered into Reference Manager for Windows, a bibliographic management program.
    We critically analyzed the literature in order to draw conclusions about our current knowledge of cost-effectiveness and cost-benefit studies of preventive interventions. We used this knowledge to construct a theoretical framework to guide readers’ thinking and research about the economics of interventions to prevent drug use.
    We hope that readers will be motivated to expand their current intervention monitoring and evaluation efforts to include costs and benefits and cost-effectiveness analyses, thus contributing to a solid body of economic information about drug prevention. We also urge that economic findings about drug prevention programs and policies be presented in policy-focused articles and presentations in order to make such information accessible to a broad audience of policy makers, researchers, practitioners, and community leaders.

EXECUTIVE SUMMARY

    Over the past 3 decades, the Department of Health and Human Services (DHHS) has funded a wide range of strategies to prevent or reduce the use of tobacco, alcohol, and other drugs. Strategies have been directed at modifying characteristics of individuals (increasing drug knowledge, changing attitudes about drugs, increasing social skills, and increasing resistance to social influence or peer pressure) and the environmental context of individuals (improving parenting skills and changing school, workplace, and community policies).
    Policy makers in governmental bodies, schools, community-based organizations, and funding agencies are being asked to justify that the benefits of investing in drug abuse prevention programs are worth the costs. They are also being asked to choose between alternate programs that seek to achieve similar goals, such as school-based versus family-based drug prevention programs. Although cost issues generally are important to policy makers, they are particularly important in an era of fiscal constraints and declining resources.
    This literature review seeks to enhance the capacity of drug prevention initiatives to answer questions about costs and benefits. A comprehensive review is provided of the literature on cost benefit-analysis (CBA) and cost-effectiveness analysis (CEA). A theoretical framework is constructed to guide thinking and research on costs and benefits and cost effectiveness. Finally, steps are recommended that will lead to answers about whether the benefits of drug prevention outweigh the costs and which interventions to prevent drug abuse are the most cost-effective.

INTRODUCTION AND BACKGROUND

Extent and Magnitude of Drug Use

    Drug use is a major public health problem because of the proportion of the population using drugs and because of the multiple effects of drug abuse on users, their families, and their communities.¹  More than 38% of adults have used marijuana or other illicit drugs at least once; a significant majority have smoked tobacco; and almost all have consumed alcohol (National Household Survey on Drug Abuse, 1996). High school seniors report that 63.2% have been drunk, 41.7% have used marijuana, and 54% have used cocaine (Johnston, O’Malley, & Bachman, 1996). High school senior data are considered lower-bound prevalence data because the sample does not include an estimated 20% who dropped out of school, a subset with higher rates of drug use than the students surveyed. Particularly disturbing is the proportion of drug users with preadolescent onset of use, with 55% of students reporting that they used alcohol by sixth grade (Johnston, O’Malley, & Bachman, 1989).
    The prevalence of drug use is disproportionately higher in economically disadvantaged subgroups of the population (National Household Survey on Drug Abuse, 1996), indicating the importance of developing or adapting drug abuse efforts for high-risk populations. In addition, although overall use has decreased over time, heavy use has remained constant (National Household Survey on Drug Abuse, 1996), indicating the importance of developing effective interventions to prevent or reduce the heavy or frequent consumption of tobacco, alcohol, or other drugs.

Impact of Drug Use

    Almost one-third of all deaths in the United States are attributed to drug use. Drug use is linked to increases in the user's health-related risk behaviors such as failure to use condoms, failure to use birth control, and sharing of intravenous needles (Cahalan, 1991). It also is linked to an increase in the risk for a number of health conditions including cancer, chronic liver disease, pancreatitis, peptic ulcers, and tuberculosis (Rice, Kelman, Miller, & Dunmeyer, 1990). Drug use has a psychological and economic impact on the user's family (Brook, Brook, Gordon, Whiteman, & Cohen, 1990) and increases the risk of infant mortality and morbidity for the childbearing user's offspring (Chasnoff, 1988; Kleinman, Pierre, Madaus, Land, & Schwann, 1988; Little, Snell, Klein, & Gilstrap, 1989). Community impacts stem from the association of drug use with motor vehicle crashes, suicide, homicide, drownings, boating deaths, rape, assault, and robbery (Grossman, Chaloupka, Saffer, & Laixuthai, 1994; Inciardi & Pottieger, 1991; Perrine, Peck, & Fell, 1988). Current analyses of the costs generated by drug use problems in the U.S. population estimate that the U.S. economy absorbed $70.3 billion in alcohol costs and $44.1 billion in drug abuse costs (Rice et al., 1990). Most of the costs of drug abuse are due to crime, including the costs associated with police protection, private legal defense, property destruction, and productivity losses for those who engage in drug-related crime or for people incarcerated in prison as a result of a drug-related crime (Rice, Kelman, & Miller, 1991). Additionally, researchers have linked substance use during high school and young adulthood to lower educational attainment and lower earnings (Cook & Moore, 1993; Kenkel & Ribar, 1994; Yamada, Kendix, & Yamada, 1996).

Common Approaches to Preventing Drug Use

    Over the past 3 decades, a wide range of prevention strategies have been directed at modifying individual characteristics (increasing drug knowledge, changing attitudes about drugs, increasing social skills, and resisting social influence or peer pressure) and the environmental context of individuals (improving parenting skills) and changing school, workplace, and community policies.²  In general, we have less information concerning the effectiveness of prevention efforts directed at modifying the environmental context of individuals, particularly interventions that modify parent and community influence. A critical review of the prevention efforts that have been evaluated reveals that the most powerful prevention programs are based on social learning models directed at behaviors linked with drug use. Social learning-based drug prevention programs have positive long-term effects on tobacco, alcohol, and marijuana use (Botvin, Baker, Dusenbury, Botvin, & Diaz, 1995; Eggert, Thompson, Herting, Nicholas, & Dicker, 1994; O’Donnell, Hawkins, Catalano, Abbott, & Day, 1995; Pentz et al., 1989). In addition, these programs produce similar improvements in drug-related behavior, such as reductions in antisocial behavior and school behavior problems, and in areas of social functioning, such as improvements in academic skills, greater commitment to school, improved drug refusal skills, and reduced affiliation with deviant peers (O’Donnell et al., 1995; Spoth, Redmond, Haggerty, & Ward, 1995). It appears that effects are stronger with repeated booster sessions at times of developmental transitions (Bell, Ellickson, & Harrison, 1993) and when school-based efforts are combined with a parenting component (Rohrbach et al., 1994).

Economic Evaluations of Drug Abuse Preventive Interventions

    Very few researchers have published rigorous economic evaluations of drug abuse prevention programs. Recently, a few researchers have addressed conceptual issues involved in the application of economic evaluation methodology to drug abuse prevention. Additionally, limited information on the costs of drug abuse prevention programs is now available. Nevertheless, at this point, very little is known about the costs and benefits of drug use prevention.
    The best document describing costs is, perhaps, Making the Grade: A Guide to School Drug Prevention Programs, which offers curriculum cost information for over 40 school-based, drug abuse preventive interventions. The curriculum cost is calculated on a per-class, per-year basis, including only the costs of supplies and materials such as videos, textbooks, workbooks, games, and lesson plans (Drug Strategies, 1996).
    For comprehensive health programs, curriculum costs per class, per year ranged from $65 for an elementary school, 10-module health course to $739-$1,527 for an integrated, elementary school science-based curriculum designed to be offered three times a week. Drug prevention programs aimed at all grade levels were somewhat lower in cost: the most expensive program cost $1,000 per class, per year but several curricula could be purchased for less than $200 per class, per year. Elementary and middle school prevention programs were fairly affordable, ranging from $24 to $465 per class, per year. Similarly, high school prevention program costs varied from $13 to $310 per class, per year (Drug Strategies, 1996).
    Tricker and Davis (1988) did a more comprehensive calculation of the total cost and the average cost per teacher hour of two drug abuse prevention programs, Here's Looking at You II and Starting Early. These programs involved three school districts in Oregon — a total of 4,325 students, 171 teachers, and 21 schools.
    Aware that teacher time may be a very important component of costs, the authors asked 44 randomly selected teachers to complete questionnaires on teacher preparation time. Other cost data were collected from program coordinators and the school districts. Costs included supplies, time, inservice training, guest speakers, coordinators' and counselors' salaries, teacher substitutes, demonstration teachers, and teachers' tuition reimbursement (Tricker & Davis, 1988). It turned out that teacher training was a large component of the program costs.
    Tricker and Davis discovered that costs differed by program and by school district. The cost per teacher hour of the Here's Looking at You II program ranged from $22 to $45 in the two school districts that implemented the program. The Starting Early program cost less per teacher hour compared to the Here’s Looking at You II program. The researchers concluded that variation in the cost of the same program across school districts may be related to district differences in teacher training, teacher preparation time, and length of the program implementation (Tricker & Davis, 1988).
    Unfortunately, this level of detailed analysis of costs is not available for other drug abuse prevention programs. Even less research is available on the benefits of drug abuse prevention programs.
    Wylie (1983) demonstrated the applicability of economic evaluation techniques to health education programs by analyzing a hypothetical school-based health education program. The author made assumptions about the costs of such a program by estimating teachers' salaries, fringe benefits, and materials (Wylie, 1983). It is not clear, however, how Wylie arrived at these estimates. The author also made assumptions about the potential benefits of the program and attempted to quantify those benefits in monetary terms. He included as benefits the prevention of smoking, alcohol-related accidents, venereal disease, unwanted pregnancy, dental caries, anxiety, influenza, and stress. Again, it is not clear how the author estimated the monetary values of the benefits of the hypothetical program or how he established causality between the health education program and the outcomes.
    Mitchel, Hu, McDonnell, and Swisher (1984) presented results from a cost-effectiveness evaluation of a drug abuse prevention program based in a parochial high school. The educational drug resistance program was created by a private, nonprofit organization and was integrated into the school's health, religion, and social studies curriculum. Students attended all, some, or none of the health, religion, and social studies components of the program.
    The costs of the program, as measured, included curriculum development, teacher training, administrative costs, program promotion, materials, and instruction. The total cost of the program was estimated to be $68 per student (Mitchel et al., 1984).
    For these high school students, the benefits of the program included lower rates of smoking and alcohol use after 5 months and lower rates of marijuana use after 9 months. These benefits, however, were limited to students who participated in the religion component of the program. Overall, a cost of $68 per student was associated with the following short-term benefits in a population of 150 students: (a) reduction of cigarette use by three packs after 3 months; (b) reduction in bottles of beer consumed by 16 bottles after 3 months; and (c) reduction in glasses of wine consumed by 10 glasses after 3 months.
    Stein, Swisher, Hu, and McDonnell (1984) conducted a cost-effectiveness analysis of a program designed to provide students with meaningful alternatives to drug use. The program, called Channel One, organized groups of children and adults to participate in community service, educational, and recreational projects. The authors found that although the average program cost per student ranged from $180 to $214, the program was not effective in reducing participants' drug use. The Channel One program therefore was not effective and, obviously, not cost-effective.
    In a more recent article, Plotnick (1994) addresses methodological issues in applying economic evaluation methods to drug abuse prevention programs. Like Weisbrod (1983), Plotnick points out that when measuring the benefits of a drug abuse prevention program, it is important to recognize and include nonmonetary as well as monetary benefits (p. 344). He provides an example of a hypothetical drug abuse prevention program's benefits. Some of the benefits, such as higher earnings and reductions in social services use, can be quantified in monetary terms. Other benefits of the hypothetical program, such as improved mental health and less family conflict, cannot be captured fully by monetary terms, but they are still an important part of the analysis (p. 355).

Economic Evaluations of Other Health and Social Preventive Interventions

    The number of published, health care-related economic evaluations has increased dramatically over the last 30 years. In a literature review of health care cost-benefit and cost-effectiveness analyses, Elixhauser, Luce, Taylor, and Reblando (1993) cited 3,206 articles published between 1979 and 1990. Economic evaluations of preventive interventions made up about 20% of these studies. Although none of these economic evaluations focused specifically on drug abuse prevention programs, some do provide a useful framework for measuring program costs and benefits.
    Preventive interventions can be classified according to the manner in which the strategies are implemented. Teutsch (1992) defines three broad prevention technologies: (a) clinical prevention strategies, (b) behavioral prevention strategies, and (c) environmental prevention strategies. Of these three approaches, evaluations of behavioral strategies are most pertinent to the study of drug abuse prevention programs. For example, the Elixhauser review includes studies that examine the costs and benefits of AIDS prevention programs, stress management programs, fertility planning programs, and patient education programs (Elixhauser, Luce, et al., 1993).
    What items should be used when calculating program costs? Gorsky, MacGowan, Swanson, and DelGado (1995) presented a cost analysis of three drug abuse treatment centers’ HIV testing/counseling programs. This study was the first to offer a program-based, instead of budget-based, cost analysis of such programs. In estimating program costs, the authors included administrative support, provider time, fringe benefits, telephone, travel, laboratory costs, and coordinator salaries. Not included are costs borne by patients (i.e., travel costs) or capital costs (i.e., the cost of space).
    Byers et al. (1995) measured the cost of a work site nutritional education program that accompanied on-the-job cholesterol screening. Their cost estimates included supplies plus the value of the participants’ lost work hours, because this program was offered to employees during their work day.
    Barnett (1985, 1993) estimated program costs from the Perry Preschool Program, an intensive early education project designed to help disadvantaged African American children. This program provided an active learning curriculum called the High/Scope Curriculum, described by Schweinhart, Barnes, and Weikart (1993). In his cost analysis, Barnett considered teacher salaries, social security, retirement, depreciation, supplies, and the potential time costs of participants and their families.
    These three cost analyses concerned programs at health clinics, work sites, and schools. The Barnett study with school children is especially useful to the topic of drug abuse prevention for youth, because prevention programs are often school-based, pull-out programs. Where such prevention programs are provided in a hospital or clinic setting, the Scheffler, Feuchtbaum, and Phibbs (1992) and Windsor et al. (1993) studies of behavioral health programs for pregnant women may provide helpful frameworks for estimating costs.
    What items should be used in calculating the value of program benefits or outcomes of a behavioral preventive intervention? Many studies in the health arena value outcomes based on prevented lost wages or hospital costs. For example, when describing a hypothetical preconception care program for pregnant women with diabetes, Elixhauser, Weschler, et al. (1993) consider the cost of poor birth outcomes and poor child health. In examining a back injury prevention program, Shi (1993) evaluates the benefits of prevented sick days and changes in back injury risk status. Hatziandreu, Koplan, Weinstein, Caspersen, and Warner (1988) study coronary heart disease morbidity and mortality that can be prevented by regular exercise.
    In drug abuse prevention programs, however, the participants are most likely young people who are not involved in the labor market. Furthermore, the outcomes of interest may be intrinsically difficult to quantify in monetary terms. Reduced school absenteeism and improved self-esteem, for example, are harder to monetize than the value of lost wages.
    Weisbrod (1983) points out that excluding program benefits that are difficult to quantify may underestimate the value of an intervention. Therefore, he suggests listing these benefits in the analysis so that policy makers can make personal judgments as to how much society values the qualitatively measured outcomes. In his classic economic evaluation of a community-based treatment program for the severely mentally ill, Weisbrod includes some benefits that are difficult to monetize, such as improved consumer decision making and improved mental health status.
    In a remarkable long-term study of the Perry Preschool Program, Barnett (1985, 1993) was able to follow participants from preschool through age 27. Ultimately, in the 1993 report, he was able to monetize program benefits as increased future earnings and fringe benefits, reduced welfare dependency, and reduced costs associated with crime. His 1985 report describing the same subjects at age 19 included data on participants’ education and juvenile crime. For these adolescents, in 1985, he quantified outcomes by estimating the costs of special education plus the victim and taxpayer costs of juvenile crime. He also projected the effects of preschool participation on future earnings and future welfare dependency (Barnett, 1985).
    In regard to the economic analysis of drug abuse prevention programs with youth, the Weisbrod and Barnett studies suggest (a) comprehensively measuring all benefits, including outcomes that are monetary, quantitative, and qualitative; (b) monetizing benefits where possible, such as school placement and juvenile crime; and (c) using available adolescent information to forecast future monetary benefits.

Economic Evaluations of Drug Abuse Treatment

    Although there are few published economic evaluations of drug abuse treatment programs, cost-benefit and cost-effectiveness analyses are not completely new in the drug abuse treatment field. Twenty years ago, researchers already had started to assess the costs and benefits of alcoholism treatment. Swint and Nelson (1977), for example, describe economic analysis tools and demonstrate their applicability to hypothetical alcoholism treatment programs. Hertzman and Montague (1977) address methodological issues in applying cost-benefit analysis to alcoholism treatment. During the 1980s, Holder and Blose (1987), Longabaugh et al. (1983), Scheffler et al. (1992), and others broadened the scope of this literature by addressing alcoholism treatment's cost offset effect and the costs and benefits of different treatment options. A comprehensive discussion of this body of work is outside the scope of this literature review. Even so, it is useful to summarize the results of a few recent studies in this area.
    Bradley, French, and Rachal (1994) report results from a comprehensive cost analysis of three methadone treatment programs. They found that the average cost per patient per year ranged from $3,750 to $4,400. Harwood, Hubbard, Collins, and Rachal (1988) use a longitudinal survey of drug users in the Treatment Outcome Prospective Study (TOPS) to analyze the economic benefits of drug abuse treatment. The researchers report that after treatment, drug users impose lower rates of crime-related costs on society, but they do not earn significantly higher wages (Harwood et al., 1988).
    Using published data on the costs of various forms of treatment, Rice et al. (1990) present cost-benefit ratios for residential and outpatient drug-free treatment programs. Residential programs appear to have quite favorable cost-benefit ratios, with measured benefits being about twice as large as measured costs (Harwood et al., 1988). Other researchers using the TOPS data have found that drug users in treatment eventually achieve better labor market outcomes, indicating that these cost-benefit ratios probably are conservative estimates (French & Zarkin, 1992).
    Holder, Longabaugh, Miller, and Rubonis (1991) compile data on effectiveness and costs of a wide variety of alcohol treatment modalities ranging from Alcoholics Anonymous to drug therapies to hypnosis. They classify each type of treatment by cost and by evidence of effectiveness. The authors find that several treatment options such as behavioral self-control raining and stress management are both inexpensive and effective. Finney and Monahan (1996), building on the methodology of Holder et al., also find stress management training to be cost-effective but, in general, they find less variation in treatment effectiveness compared to Holder et al. (1991).
    The literature on the economic evaluation of drug abuse treatment also contributes important information concerning methodological issues involved in economic evaluation, and the links between economic evaluation and policy analysis and formulation.
    Apsler (1991) discusses the importance of developing and using more rigorous methods to test the effectiveness of drug treatment. He argues that our understanding of the cost-effectiveness of drug abuse treatment is limited because of the lack of cost-effectiveness studies and because of methodological weaknesses with existing studies, including reliance on single-system designs and unverified self-reports of drug use and drug-related behavior. Kim, Crutchfield, Williams, and Hepler (1994) propose that the threshold-gating strategy be used to increase the methodological rigor of costs and benefits evaluation in single-system designs. Hubbard and French (1991) propose that investigating treatment careers, components of treatment, and the moderating effect of client impairment will give us a better understanding about ways to improve effectiveness and to increase cost-benefit ratios.
    Harwood (1991) emphasizes the advantages of using cost-benefit analysis as a tool for developing a comprehensive national strategy to prevent and reduce drug abuse. He suggests using cost-benefit analysis to shape the respective roles of federal, state, and local governments in the national drug control strategy, and to identify the optimal balance between supply reduction activities (law enforcement) and demand reduction activities (prevention and treatment). Hser and Anglin (1991) highlight the importance of conducting longitudinal cost-effectiveness analyses to forecast long-term outcomes and simulate outcomes of alternative policies for planning purposes. Sindelar (1991) urges economic evaluations to investigate expected marginal benefit (costs averted) in order to plan and prioritize programs.
    These studies are useful because they address the substantive and methodological issues involved in finding the most cost-effective treatment modality, and they highlight the critical role of economic evaluations in analyzing and formulating drug abuse policy. Furthermore, many of the costs and benefits addressed in the drug abuse treatment literature are similar to those that are relevant to an economic evaluation of a drug abuse prevention program. For example, both drug abuse treatment programs and drug abuse prevention programs may reduce future criminality and increase future wages. At this point, however, researchers have not considered these types of long-term benefits in an evaluation of a drug abuse prevention program.

Conclusions from the Literature

    The three areas of economic evaluation reviewed above offer some guidance as to what types of costs and benefits might be included in the economic evaluation of a drug abuse prevention program. As suggested in Table 1, costs could include all supplies and materials used by the program, capital costs (i.e., computers, buildings), administrative costs, donated goods, volunteer labor, staff salaries and fringe benefits, staff training, and costs incurred by participants and their families. Benefits might include improvements in mental health and behavior, physical health benefits, lower rates of criminality, reduced services use, greater school and labor market productivity, and family benefits such as improved children’s outcomes and lower rates of family conflict. The range and types of costs and benefits included will vary widely according to the study design and the intervention of interest.

THEORETICAL FRAMEWORK

Utility of Cost-Effectiveness and Cost-Benefit Data for Designing and Refining Health and Human Services

    Policy makers, researchers, and practitioners seek to design and improve services to prevent the use of tobacco, alcohol, and other drugs. Reforming existing systems of care involves an implicit or explicit comparison between two or more options. Economic evaluations of specific prevention programs and policies can contribute unique and meaningful information to aid in decisions concerning modification of existing services and development of new services.
    Theoretically, then, economic evaluation methods can be very useful tools in decision making. But how are these methods currently being used to make decisions in practice? At this time, cost-effectiveness analysis methods are influential in decision making in the pharmaceutical industry. Outside this industry, despite the growing economic evaluation literature, formal economic evaluations are not commonly used in other health program decision-making processes (Russell et al., 1996). The theoretical framework presented below may be used to guide research on costs and benefits and cost-effectiveness of drug abuse preventive interventions, thus contributing valuable information for decision making and planning of health and human services.

Cost-Effectiveness Analysis Versus Cost-Benefit Analysis

    Drummond, Stoddart, and Torrance (1987) define economic evaluation as "the comparative analysis of alternative courses of action in terms of both their costs and consequences." The heart of this process is the concept of opportunity cost, in which the true cost of a drug abuse preventive intervention is essentially the foregone benefits that could have been achieved had the resources been used for the next best alternative. For example, the cost of a drug abuse prevention program that prevents a thousand children from using drugs may be a year of life of an elderly person, whose life could have been prolonged if the resources had been allocated toward an experimental therapy. When policy makers allocate funds toward a particular program, they essentially are deciding that society will give up the benefits of some other program. Economic evaluation can help decision makers allocate resources, while also attempting to ensure that limited funds are used efficiently.
    This notion of an opportunity cost is particularly important when a health program is the focus of the analysis. Unlike other parts of the economy, many goods produced in the health sector are not explicitly bought and sold in markets. Normally, a market price reflects how much a society is willing to pay for a certain good or service. For example, according to economic theory, teachers' salaries indicate how much society values the education of its children. However, the amount society is willing to pay to prevent one child from using drugs is yet to be defined. It is difficult to answer this question because prevention cannot be bought and sold in a market. This problem makes it particularly important that the opportunity costs of health interventions be made explicit; otherwise, the lack of prices to guide decision makers impedes efficient resource allocation.
    The most common economic methods used to evaluate programs are cost-effectiveness analysis and cost-benefit analysis.³  These two methods are used to compare the costs and the outcomes of alternative, competing programs. Cost-benefit analysis and cost-effectiveness analysis are similar in the methods used to collect data on costs. Both require documentation of the total value of resources consumed by the program under evaluation as well as other alternative programs being investigated. The methods, however, diverge in their treatment of the consequences, or the benefits, of the program and its alternative(s).

Cost-Effectiveness Analysis

    Cost-effectiveness analysis is implemented under the assumption that the program under evaluation and its alternative both produce the same type of outcomes. For example, the analysis may compare two drug abuse prevention programs (a parent training program versus a family training program). Or, the analysis could compare a defined drug abuse prevention program, such as a school-based social skills intervention, with the school's usual efforts with high-risk youth. The value of these outcomes themselves is not questioned; instead, the evaluator is interested in the less expensive means of producing these outcomes.
    In cost-effectiveness analysis, the question of interest is, "Which of the available alternatives is the least expensive way to produce a unit of drug use prevention?" Units of prevention can be measured in a variety of ways (e.g., life-years gained, hospital emergency room visits prevented, cases of adolescent drug use prevented), but they must be measured the same way across alternatives.

Cost-Benefit Analysis

    Cost-benefit analysis theoretically can be used to assess whether a program or policy intervention is a worthwhile investment in and of itself, without comparison to other programs. It also can be used to compare interventions and policies. Traditionally, benefits as well as costs are valued in monetary terms. This feature distinguishes cost-benefit analysis from cost-effectiveness analysis in which benefits are measured in their natural units. Cost-benefit analysis is used to determine whether the benefits of a program measured in dollars outweigh its costs and thus justify the allocation of resources to that program. The most common indices in cost-benefit analysis are the cost-benefit ratio and net benefits.

Approaches to Valuation

    The choice of approach in the valuation of costs and benefits in cost-benefit analysis reflects the researcher’s assumptions and values. The willingness-to-pay approach attempts to capture what individuals would be willing to pay for costs and benefits. For example, if an intervention reduces the probability of death or illness, willingness-to-pay methods would attempt to find what people would be willing to pay for a reduction in the probability of illness or death.
    Willingness to pay for health outcomes is difficult to measure accurately for a number of reasons. For example, individuals' willingness to pay for a health improvement may be influenced by income level (i.e., upper income families are able to pay more than poor families). Furthermore, individuals are not accustomed to placing an explicit value on probability of illness or death. There is a growing literature on willingness-to-pay methods in the environmental and health economics literatures.
    The human capital approach appears more appropriate for an assessment of the costs and benefits of drug abuse prevention because of current limitations in accurately measuring the willingness to pay for health outcomes. Under this approach, an individual's worth is measured by the discounted value of the individual's stream of productivity over time as measured by wages. The human capital approach assumes a societal perspective (discussed below) and uses data that are more readily available and reliable. The human capital approach is appropriate for determining the economic cost of a disease or condition for a defined time period or for determining the cost savings of a specific procedure or intervention.
    This approach, however, is limited when evaluating programs involving children or socially or economically disadvantaged individuals, since society tends to value its members for reasons unrelated to their productive capacity. Also, because of its dependence on market earnings, the human capital approach tends to undervalue certain other factors, such as pain and suffering (Rice et al., 1990). The ethical implications of the human capital method and economic evaluation in general are addressed below.

Classifying Costs and Benefits

    Direct and indirect benefits are classified under the more general category of core benefits. Core benefits are typically those that result directly from preventing the illness or condition itself. Core benefits include direct costs avoided, such as dollar expenditures on health, mental health, and social services related to drug misuse, and indirect costs avoided, which include the value of lost/reduced productivity. For example, if a patient participates in an inpatient drug abuse treatment program, the hospital expenses incurred are direct costs, whereas the wages lost by the patient are indirect costs. If this case of drug abuse had been prevented, the foregone hospital expenses and lost wages could be classified respectively as the direct and indirect benefits of drug abuse prevention, as suggested in Table 2.
    Other related costs are secondary to the condition under study, pertaining instead to the non-health effects of the illness. Like core benefits, other related benefits include direct costs avoided, for which monetary payments are actually made, and indirect costs avoided, which represent lost resources. Other related benefits include direct benefits such as dollar expenditures avoided on drug abuse-related services, and indirect benefits such as the value of delinquency or criminal activity avoided.

Methods for Quantifying Intangible Costs and Benefits

    Once the data have been collected, values must be assigned to costs, and in the case of cost-benefit analysis, to benefits. Normally, most costs already will be measured in dollar terms. An important issue that arises in most economic evaluations is that some costs and benefits may be difficult to value in monetary terms. Other costs and benefits even may be difficult to describe. For example, a treatment intervention may cause physical pain or anxiety. These factors are intangible costs of the intervention, and they may be difficult to describe and impossible to value accurately in dollars. This problem also arises in the context of benefits. For example, a school-based intervention may help children earn higher grades. This benefit may be easy to describe, but it is still difficult to value in dollars. Researchers have attempted to quantify intangible costs and benefits using a variety of innovative methods. The "cost" of physical pain, for example, can be estimated by a patient's expenditure on pain medication (Drummond et al., 1987). Questionnaires and experiments based on the willingness-to-pay approach can be used to elicit values for intangible costs and benefits. Drummond et al. (1987) point out that it is important to assess whether using these relatively new methods to value intangible factors truly will aid decision making. If not, it may be better to avoid this often difficult and expensive process (see Table 2).

Discounting and Sensitivity Analysis

    It is also very important to adjust future costs and benefits through a procedure called discounting. The purpose and process of discounting is best described with an example. Most people are familiar with the concept of gaining interest on an investment. Assume the interest rate is 10%. After a year in the bank, a $100 savings account will be worth $110.⁴  Discounting reverses this relationship — this reversal implies that $110 received a year from now is worth only $100 today (Banta & Luce, 1983). The present value of $110 received next year is $100.
    In other words, discounting accounts for the fact that $100 received now is worth more than $100 received a year from now, because money received now can earn interest in the bank for a year. Discounting should be performed if benefits and/or costs occur more than 1 or 2 years into the future. For preventive interventions, benefits are often realized far into the future. Because these benefits are heavily discounted, they may appear to be worth very little.
    The issue of discounting becomes controversial when the choice of interest rate is disputed. Usually, a rate of 2-10% is considered to be consistent with economic theory — 5% is a commonly used rate (Drummond et al., 1987). Most recently, however, an expert panel organized by the United States Public Health Service recommended that researchers use a baseline 3% discount rate (Lipscomb, Weinstein, & Torrance, 1996).
    Often, the evaluator will try a range of rates to assess the implications of "worst-case" and "best-case" scenarios (Banta & Luce, 1983). This process is called sensitivity analysis and is now considered to be an essential element of a cost-benefit or cost-effectiveness evaluation (Drummond et al., 1987). Sensitivity analysis is not limited to analysis of the interest rate. Any uncertain assumptions or figures should be assigned different values to see whether these changes affect the results or the conclusions of the study. If the results or conclusions of the study are the same over a range of values and assumptions, the evaluator can make a recommendation with a degree of confidence. If not, the evaluator should list the range of values or assumptions that correspond to a specific result (Drummond et al., 1987).

Cost-Effectiveness Ratios and Net Benefits

    Cost-effectiveness ratios can be compared across interventions that have the same outcomes. For example, if drug abuse prevention program A costs $10 per case of drug abuse prevented, and drug abuse prevention program B costs $15 per case of drug abuse prevented, clearly program A is a less costly way of achieving the same outcome.
    Cost-benefit ratios and net benefits figures allow the evaluator both to assess the economic worth of the program and to compare the program to competing alternatives. For example, if program A has $100 of net benefits and program B has $200 of net benefits, both programs are worth implementing because their valued benefits exceed their costs. If only one program can be selected, program B is preferable because it maximizes net benefits.

Ethical Implications of Economic Evaluations

    Surely, cost-benefit and cost-effectiveness analyses will become more commonplace in decision making as researchers’ and policy makers’ familiarity with the methods grows. Even so, it is important to acknowledge that many people are uncomfortable with some of the ethical implications of economic evaluation. This discomfort is especially acute when economic evaluation is applied to the health sector.
    Economic evaluation methods focus on efficiency, not equity. Theoretically, equity can be achieved through redistribution after an efficient outcome is reached. In practice, however, redistribution is often difficult to implement. For example, an efficient allocation might result in gains for wealthy people and losses for indigent people. If one of the goals of the health sector is to promote greater equity, the efficient outcome will not be appropriate.
    Economic evaluation methods are not value free, and ethical issues often arise in their application (Drummond et al., 1987). But decision makers face ethical dilemmas and make implicit value judgments when economic evaluation methods are not used. The advantage of economic evaluation methods is that values and judgments are made explicit. Furthermore, the results of an economic evaluation alone should never be used to make a decision. Instead, decision making should rely on cost-benefit and cost-effectiveness analyses combined with other analyses and other considerations. Russell et al. (1996) compare the role of cost-effectiveness analysis to a Consumer Reports article. Decision makers use Consumer Reports for useful information about a product, but they usually consider many other factors before making a decision (Russell, 1996).

APPLICATION

    Many issues arise in the application of economic evaluation methodology to policies, programs, and interventions. Economic evaluations come in a wide variety of forms, and every evaluation project will have unique challenges. For this reason, it is not possible to write an "economic evaluation cookbook" or a standardized economic evaluation computer program. Even so, it may be useful to offer some general, practical guidelines to the economic evaluation process.

    The objective of the study may lead to a natural choice of a method. For example, if a health department wants to compare two programs with different outcomes, CEA will not be an option. But in some cases, several methods of analysis may be possible, and the choice of method will depend on a variety of factors. One important factor to be considered is whether the objective of the evaluation is to compare the worth of the intervention’s objective to the worth of an alternative’s objectives. If so, CBA will be more appropriate than CEA, which assumes from the start that the intervention’s objective is worth pursuing.
    In many cases, practical considerations will dominate the choice of methods. For example, even if a CBA is desirable, it may be impossible or very difficult to value all of the outcomes of a particular intervention in monetary terms. Usually, a CEA is the "easier" choice when the outcomes are difficult to value in monetary terms. Whether the methods selected are a natural choice or a difficult decision, it is useful to review any existing literature that addresses problems similar to the one under evaluation. Often, the evaluator will be able to reassess and/or refine the choice of methods after reading about problems others have faced in conducting a CBA or CEA with similar goals.
    Many of the solutions to economic evaluation problems require a thorough understanding of the intervention, its alternative(s), and the environment(s) in which they exist. The evaluator also must describe completely both the intervention under evaluation and its competing alternative(s). The objectives of the intervention and its alternative(s) should be clearly stated. This step is extremely important in part because it highlights the fact that resources used to implement the intervention under evaluation could have been used elsewhere. CBA theoretically can be used to determine the worth of an intervention without explicit comparison to an alternative because the benefits and costs are both measured in dollars and can be compared. In most practical cases, however, alternatives exist, and even if the alternative is "do nothing," it must be documented.

In the Application of Economic Evaluation Methodology, It Is Important that the Evaluator Clearly Define the Perspective of the Study

    It is important to specify whether costs and consequences are viewed as accruing to private firms and individuals or to society as a whole. Often, the private viewpoint is too restrictive. For example, a health program's costs might outweigh its benefits from the perspective of a single hospital that provides the program. But if societal benefits outweigh societal costs, efficiency would be enhanced if resources were allocated toward the program so that the hospital or some other institution was willing to provide the program (Drummond et al., 1987).⁵  In most cases, the societal viewpoint is most suitable, particularly for health care evaluations that involve public dollars (Torrance, Siegel, & Luce, 1996). The perspective of the study will determine which costs and benefits are included in the analysis. It makes sense, therefore, to decide on an appropriate perspective at the outset of the evaluation.
    After choosing between CEA and CBA, and choosing a study perspective, policy makers, researchers, or practitioners may move through five general steps that make up the economic evaluation process. These general steps are presented as a series, but in some cases it may be useful to change the order.

    The first step in the design of an economic evaluation is to define the question of interest comprehensively. The evaluators should state clearly (a) the program under evaluation and its alternative; (b) the perspective of the study; and (c) the types of economic evaluation that are appropriate and feasible. Because many important decisions about the analysis will be made at this early stage, it is often useful to convene an interdisciplinary economic evaluation work group at the beginning of the evaluation process.

    After the question is well defined, it is important to have a practical evaluation plan. The plan should include a framework for collecting cost and outcome data from the programs. In many cases, existing data collection systems may be inadequate for an economic evaluation. For example, accounting records may be available that document the costs of a new drug abuse prevention program and its competing alternative. Accounting records, however, usually do not include many of the opportunity costs of programs such as volunteer labor and shared office space.
    In many cases, the evaluator will be able to build on existing data collection instruments and existing sources of data on costs and outcomes. Even so, it is very important to study the existing data sources carefully and to work closely with the programs’ staff to ensure that the methods used to collect data are consistent with economic evaluation methodology.

    Economic evaluation methods force society to place an explicit value on costs and benefits. Often, this procedure has ethical implications. For example, if increased wages were used to quantify the benefits of a drug abuse prevention program, we would place very low values on the outcomes of children, elderly people, and disabled people. Naturally, this valuation method is likely to clash with the goals of evaluators and society as a whole. For this reason, the economic evaluation work group should discuss the ethical implications of the analysis early in the evaluation process.

    After collecting and appropriately measuring data on costs and outcomes, the evaluators can calculate cost-effectiveness ratios, cost-benefit ratios, and/or net benefits figures for the program(s) under study. As described in the previous section, discounting and sensitivity analysis should be used at this stage to adjust costs and benefits for differential timing and uncertainty.

    Economic evaluation methods make assumptions, values, and judgments explicit and can be combined with other analyses for effective decision making. For example, the human capital approach described earlier essentially values human beings by their expected lifetime earnings. One problem with this approach is that elderly people would be assigned low values, since they no longer work. If this implication is not stated in the analysis, users of the study results might misinterpret the findings and make undesirable policy decisions. Banta and Luce (1983) point out that "quantitative results are powerful and may overwhelm the policy maker with a false sense of security" (p. 154). It is the evaluator’s responsibility to prevent this possibility by uncovering and discussing any implicit assumptions that may have been made in the analysis.

FUTURE DIRECTIONS

    The objective of an economic evaluation of a preventive intervention is to answer the question, "Do the benefits of this intervention outweigh its costs?" Ideally, if a fairly large literature exists in a particular area, one also will be able to answer the question, "Are certain prevention programs more cost-effective than others?" To make informed choices, researchers, policy makers, and taxpayers need answers to both of these questions.
    At this point, however, the economic evaluation literature in the area of drug abuse prevention is very limited. It would be helpful if researchers developed a standardized methodology that could be used to create summary "cost per participant" ratios from programs’ existing records. Prevention programs also need to be evaluated routinely for short-term effectiveness, and at least a subset of programs should be evaluated for long-term effectiveness. Future research should address the following questions:

1. What do prevention programs cost?
   
2. Do outcomes justify the costs?
   
3. Is intervention cost associated with outcome (i.e., Do more expensive programs have better outcomes than less expensive programs)?
   
4. Are number of intervention components associated with outcome (i.e., Do programs with more intervention elements have better outcomes than programs with fewer intervention elements)?
   
5. Is program delivery method associated with outcome (i.e., Do interactive interventions have better outcomes than didactic interventions; do characteristics of the intervenors influence outcome)?
   
6. Is program delivery site associated with outcome (i.e., Do community-based interventions have better outcomes than school-based interventions)?

    Answering these questions involves many methodological challenges. For instance, documenting program costs may not be as simple as expected. Because accounting records generally are not kept for billing purposes, they tend to be of poorer quality than treatment records. Also, because many prevention programs are relatively new, they may not use an accounting system that sufficiently disaggregates costs. This problem is magnified by the fact that a number of cost issues cross intervention and comparison conditions. For small programs with little or no institutional support or accounting infrastructure, it is hard to document the costs of prevention services for youth as compared to a comparison group involved in the usual and customary efforts (i.e., the status quo).
    In addition, decisions must be made about handling capital costs and joint costs. For example, prevention programs often share space, administrative staff, and office equipment with other programs. An appropriate portion of these costs must be allocated to the program under evaluation. Capital equipment depreciation and the opportunity cost of capital also must be included. Furthermore, the opportunity cost of donated goods and time must be valued and included in the analysis.
    When measuring an effect, it is important to make an assumption about durability, that is, how long an effect will last. For example, if a prevention program is designed to raise self-esteem in children in an effort to prevent drug use, important questions about the outcome are "How long will it take for self-esteem to be raised to a level that is defined as success?" and "Will the effects of the increase in self-esteem last through childhood? Into adolescence? Into adulthood? Will this affect adult wages?" Unfortunately, the impact of an intervention may take years to realize, but the average study is limited to 4 years or less. This time period may not be sufficient to measure the long-term benefits of preventive interventions, such as the use of health/mental health services or the labor market experience. These outcomes can be measured and, in some cases, valued. But very few projects last long enough to follow youth into their young adult years when health/mental health and labor market outcomes can be measured.
    In fact, prevention implies that benefits are realized far into the future. For example, children who avoid drug use might complete more years of education, have fewer adult drug use problems, and earn higher wages as adults compared to those who used alcohol and illicit drugs during youth.
    The Public Health Service recommends the use of models and secondary data sources when primary data are inadequate or inappropriate (Torrance et al., 1996). For example, longitudinal data on the labor market experiences of young drug users could be used to predict the labor market benefits of drug use prevention. Estimates of the duration of program effects could be obtained from other studies. Because the modeling of long-term prevention benefits usually involves uncertainty and many assumptions, it is essential that sensitivity analyses are conducted (Gold, Siegel, Russell, & Weinstein, 1996).
    Finally, the drug abuse prevention program must be causatively linked to the outcomes of interest. If a drug abuse prevention program is not effective, it will never be cost-effective (Drummond et al., 1987). Ideally, a randomized controlled trial should be used to assess the effectiveness of the program. In many cases, however, this study design will not be feasible. Nevertheless, a wide variety of study designs and statistical methods exist that can assess causality. Mandelblatt et al. (1996) review different methods for assessing the effectiveness of health interventions.

Promoting More Economic Evaluations of Drug Prevention Programs and Policies

    In many cases, existing data collection systems for drug prevention initiatives may be used as a foundation for an economic evaluation. Policy makers, researchers, and practitioners may augment the existing system with specific data elements pertaining to costs and benefits (see Table 1). For example, accounting records may be available that document the costs of a new drug abuse prevention program and its competing alternative. But accounting records often do not include many of the opportunity costs of programs such as volunteer labor and shared office space. Similarly, existing information on outcomes might be limited to information on drug use initiation, even though many other program benefits might exist. Outcomes data also may include only very short-term outcomes. In many cases, the evaluator will be able to build on existing sources of data on costs and outcomes.
    We hope that readers will be motivated to expand their current intervention monitoring and evaluation efforts to include costs and benefits and cost-effectiveness analyses, thus contributing to a solid body of economic information about drug prevention. We also urge that economic findings about drug prevention programs and policies be presented in policy-focused articles and presentations in order to make such information accessible to a broad audience of policy makers, researchers, practitioners, and community leaders.
    An innovative way to make information about costs and benefits and cost-effectiveness of drug prevention programs and policies accessible would be to develop an electronic archive system through an Internet Web site located at the National Institute on Drug Abuse. This system would archive economic evaluation datasets that include common measures of costs and benefits from drug prevention efforts across the country. A knowledge-based graphical user interface to drug prevention economic information could be the primary means for increasing true accessibility to (and thereby use of) the available information concerning costs and benefits and cost-effectiveness of various drug prevention efforts. We envision that the primary audience for this support system would be policy users who are knowledgeable about the major issues in the field, but are not fluent enough with information systems to generate and test issue-oriented hypotheses online.

CONCLUSIONS

    With evidence that adolescent drug use has been rising in recent years, questions about the value of prevention programs are once again prominent in the public debate. Moreover, questions about program effectiveness are being increasingly linked with questions about program costs. Unfortunately, the literature on the cost-effectiveness and costs and benefits of prevention programs is relatively new and limited in scope.
    It is important that more drug abuse prevention programs undergo economic evaluation. Public policy makers are currently choosing to reduce funding for drug abuse prevention initiatives while maintaining dollars devoted to supply reduction efforts that largely rely on the use of law enforcement officials. The inability to answer, with confidence, the question of "what works" has attenuated policy makers’ support for even those drug abuse prevention efforts that have documented evidence of success.
    Although many contend that drug prevention programs are more cost-effective than treatment, there is little evidence of the financial costs and benefits of these programs. As discussed earlier, the lack of research in this area reflects two critical challenges: difficulties assigning costs to intervention and comparison conditions, and problems in measuring outcomes that do not generally occur until many years after a program’s completion.
    Policy makers want to know how to use public funds to reduce drug abuse. Although we have strong evidence about how to prevent the use of tobacco, alcohol, and other drugs, there is limited information about whether the benefits of prevention outweigh the costs, and which drug prevention efforts are most cost-effective. The theoretical framework outlined in this literature review can be used to guide thinking and research concerning the economics of drug abuse prevention. In many cases, existing intervention monitoring and evaluation schemes can be augmented with elements from the framework to provide valuable information concerning costs and benefits of drug prevention programs and policies. We urge all ongoing prevention efforts and developing prevention efforts to consider augmenting existing research with an economic evaluation of the program or policy. We are currently allocating to prevention research a small fraction of the public dollars spent on drug treatment and law enforcement. Our hope for a better future lies in the realm of prevention; to build drug abuse prevention, we must provide the critical cost information that policy makers need to make informed choices about our nation’s resources.

Table 1: Cost and Benefit Components for Drug Abuse Prevention Program Evaluations

NOTES

1. Drug use in this paper refers to the use of tobacco, alcohol, and illicit drugs.  Return to text.

2. Parenting skills refers to improved parent monitoring of child behavior and parent-child communication.  School, workplace, and community policies include laws or policies creating drug-free environments, restricting the sale and distribution of tobacco and alcohol to minors, raising the minimum drinking age, regulating tobacco and alcohol advertising, and raising the price of tobacco and alcohol.  Return to text.

3. Cost-utility analysis is another method of economic analysis, but it will not be discussed in this paper.  Return to text.

4. For simplicity, ignore issues of monthly compounding of interest.  Return to text.

5. The term efficiency as used here refers to the concept of Pareto efficiency in economics.  As allocation is Pareto efficient if no other allocation can make an individual better off without making at least one other individual worse off.  If societal benefits are greater than societal costs and the program is not implemented, the current allocation is not Pareto efficient.  Return to text.

REFERENCES

Apsler, R. (1991). Evaluating the cost-effectiveness of drug abuse treatment services. In W. S. Cartwright & J. M. Kaple (Eds.), Economic costs, cost-effectiveness, financing, and community-based drug treatment. NIDA Research Monograph 113 (pp. 57-66). (DHHS Publication No. ADM 91-1823). Washington, DC: U.S. Government Printing Office.

Banta, H. D., & Luce, B. R. (1983). Assessing the cost-effectiveness of prevention. Journal of Community Health, 9 (2), 145-165.

Barnett, W. S. (1985). The Perry Preschool Program & its long-term effects: A benefit-cost analysis. Ypsilanti, MI: High/Scope Educational Research Foundation.

Barnett, W. S. (1993). Cost-benefit analysis. In L. Schweinhart, H. V. Barnes, & D. P. Weikart (Eds.), Significant benefits: The High/Scope Perry Preschool Study through age 27 (pp. 142-171). Ypsilanti, MI: High/Scope Educational Research Foundation.

Bell, R. M., Ellickson, P. L., & Harrison, E. R. (1993). Do drug prevention effects persist into high school? How Project Alert did with ninth graders. Preventive Medicine, 22, 463-483.

Botvin, G. J., Baker, E., Dusenbury, L., Botvin, E. M., & Diaz, T. (1995). Long-term follow-up results of a randomized drug abuse prevention trial in a white middle-class population. Journal of the American Medical Association, 273 (14), 1106-1112.

Bradley, C. J., French, M. T., & Rachal, J. V. (1994). Financing and cost of standard and enhanced methadone treatment. Journal of Drug Abuse Treatment, 11 (3), 433-442.

Brook, J. S., Brook, D. W., Gordon, A. S., Whiteman, M., & Cohen, P. (1990). The psychosocial etiology of adolescent drug use: A family interactional approach. Genetic, Social, and General Psychology Monographs, 116 (2).

Byers, T., Mullis, R., Anderson, J., Dusenbury, L., Gorsky, R., Kimber, C., Krueger, K., Kuester, S., Mokdad, A., Perry, G., & Smith, C. A. (1995). The costs and effects of a nutritional education program following work-site cholesterol screening. American Journal of Public Health, 85 (5), 650-655.

Cahalan, D. (1991). An ounce of prevention: Strategies for solving tobacco, alcohol and drug problems. Washington, DC: Jossey-Bass.

Chasnoff, I. J. (1988). Drug use in pregnancy: Parameters of risk. Pediatric Clinics of North America, 35, 1403-1412.

Chatterji, P., Lillie-Blanton, M., & Werthamer, L. A. (May, 1997). Economic evaluation of drug abuse prevention programs. NIDA Literature Review.

Cook, P. J., & Moore, M. J. (1993). Drinking and schooling. Journal of Health Economics, 12, 411-429.

Drug Strategies. (1996). Making the grade: A guide to school drug prevention programs. Washington DC.

Drummond, M., Stoddart, G., & Torrance, G. (1987). Methods for the economic evaluation of health care programmes. New York: Oxford Medical Publications.

Eggert, L. L., Thompson, E. A., Herting, J. R., Nicholas, L. J., & Dicker, B. G. (1994). Preventing adolescent drug abuse and high school dropout through an intensive school-based social network development program. American Journal of Health Promotion, 8 (3), 202-215.

Elixhauser, A., Luce, B. R., Taylor, W. R., & Reblando, J. (1993). Health care cost-benefit and cost-effectiveness analysis (CBA/CEA) from 1979 to 1990: A bibliography. Medical Care, 31 (7), JS1-JS11.

Elixhauser, A., Weschler, J. M., Kitzmiller, J. L., Marks, J. S., Bennert, H. W., Coustan, D. R., Gabbe, S. G., Herman, W. H., Kaufman, R. C., Ogata, E. S., & Sepe, S. J. (1993). Cost-benefit analysis of preconception care for women with established diabetes mellitus. Diabetes Care, 16 (8), 1146-1157.

Finney, J. W., & Monahan, S. C. (1996). The cost-effectiveness of treatment for alcoholism: A second approximation. Journal of Studies on Alcohol, 57, 229-243.

French, M. T., & Zarkin, G. A. (1992). The effects of drug abuse treatment on legal and illegal earnings. Contemporary Policy Issues, 10 (2), 98-110.

Gold, M. R., Siegel, J. E., Russell, L. B., & Weinstein, M. C. (Eds.). (1996). Cost-effectiveness in health and medicine. New York: Oxford University Press.

Gorsky, R. D., MacGowan, R. J., Swanson, N. M., & DelGado, B. P. (1995). Prevention of HIV infection in drug abusers: A cost analysis. Preventive Medicine, 24, 3-8.

Grossman, M., Chaloupka, F. J., Saffer, H., & Laixuthai, A. (1994). Effects of alcohol price policy on youth: A summary of economic research. Journal of Research on Adolescence, 4 (2), 347-364.

Harwood, H. J. (1991). Economics and drugs: Promises, problems, and prospects. In W. S. Cartwright & J. M. Kaple (Eds.), Economic costs, cost-effectiveness, financing, and community-based drug treatment. NIDA Research Monograph 113 (pp. 57-66). (DHHS Publication No. ADM 91-1823). Washington, DC: U.S. Government Printing Office.

Harwood, H. J., Hubbard, R. L., Collins, J. J., & Rachal, J. V. (1988). The costs of crime and the benefits of drug abuse treatment: A cost-benefit analysis using TOPS data. In C. G. Leukefeld & F. M. Tims (Eds.), Compulsory treatment of drug abuse research and clinical practice. NIDA Research Monograph 86 (pp. 209-235). Washington DC: U.S. Government Printing Office.

Hatziandreu, E. I., Koplan, J. P., Weinstein, M. C., Caspersen, C. J., & Warner, K. E. (1988). A cost-effectiveness analysis of exercise as a health promotion activity. American Journal of Public Health, 78 (11), 1417-1421.

Hertzman, M., & Montague, B. (1977). Cost-benefit analysis and alcoholism. Journal of Studies on Alcohol, 38 (7), 1371-1385.

Holder, H. D., & Blose, J. O. (1987). Impact of changes in distilled spirits availability on apparent consumption: A time series analysis of liquor-by-the-drink. British Journal of Addiction, 82, 623-631.

Holder, H., Longabaugh, R., Miller, W. R., & Rubonis, A. V. (1991). The cost-effectiveness of treatment for alcoholism: A first approximation. Journal of Studies on Alcohol, 52 (6), 517-540.

Hser, Y., & Anglin, M. D. (1991). Cost-effectiveness of drug abuse treatment: Relevant issues and alternative longitudinal modeling approaches. In W. S. Cartwright & J. M. Kaple (Eds.), Economic costs, cost-effectiveness, financing, and community-based drug treatment. NIDA Research Monograph 113 (pp. 67-93). (DHHS Publication No. ADM 91-1823). Washington, DC: U.S. Government Printing Office.

Hubbard, R. L., & French, M. T. (1991). New perspectives on the benefit-cost and cost-effectiveness of drug abuse treatment. In W. S. Cartwright & J. M. Kaple (Eds.), Economic costs, cost-effectiveness, financing, and community-based drug treatment. NIDA Research Monograph 113 (pp. 94-113). (DHHS Publication No. ADM 91-1823). Washington, DC: U.S. Government Printing Office.

Inciardi, J., & Pottieger, A. (1991). Crime and other drugs. Journal of Drug Issues, 12 (4).

Johnston, L. D., O’Malley, P. M., & Bachman, G. P. (1989). Drug use, drinking, and smoking: National survey results from high school, college, and young adult populations, 1975-1988. National Institute on Drug Abuse. Washington, DC: Government Printing Office.

Johnston, L. D., O’Malley, P., & Bachman, G. P. (1996). National survey results on drug use from the Monitoring the Future Study 1975-1995. NIH, USDHHS PHS Volume I, Secondary School Students.

Johnston, L. D., O’Malley, P., & Bachman, G. P. (1997). National trends in drug use and related factors among American high school students and young adults, 1975-1986. Washington, DC: U.S. Government Printing Office, DHHS Publication Number (ADM) 87-1535.

Kenkel, D. S. & Ribar, D. C. (1994). Alcohol consumption and young adults’ socioeconomic status. Brookings Papers: Microeconomics, pp. 119-173.

Kim, S., Crutchfield, C., Williams, C., & Hepler, N. (1994). An innovative and unconventional approach to program evaluation in the field of substance abuse prevention: A threshold-gating approach using single system evaluation designs [Monograph]. Journal of Community Psychology, CSAP Special Issue, pp. 61- 78.

Kleinman, J. C., Pierre M. B., Madaus, J. H., Land, J. H., & Schwann, W. F. (1988). The effects of maternal smoking on fetal and infant mortality. American Journal of Epidemiology, 127, 274-282.

Lipscomb, J., Weinstein, M. C., & Torrance, G. W. (1996). Time preference. In M. R. Gold, J. E. Siegel, L. B. Russell, & M. C. Weinstein (Eds.), Cost-effectiveness in health and medicine (pp. 214-238). New York: Oxford University Press.

Little, B. B., Snell, L. M., Klein, V. R., & Gilstrap, L. C., III. (1989). Cocaine abuse during pregnancy: Maternal and fetal implications. Obstetrics and Gynecology, 73, 157-160.

Longabaugh, R., McCrady, B., Fink, E., Stout, R., McAuley, T., Doyle, C., & McNeill, D. (1983). Cost effectiveness of alcoholism treatment in partial vs. inpatient settings. Journal of Studies on Alcohol, 44 (6), 1049-1071.

Mandelblatt, J. S., Fryback, D. G., Weinstein, M. C., Russell, L. B., Gold, M. R., & Hadorn, D. C. (1996). Assessing the effectiveness of health interventions. In M. R. Gold, J. E. Siegel, L. B. Russell, & M. C. Weinstein (Eds.), Cost-effectiveness in health and medicine (pp. 135-175). New York: Oxford University Press.

Mitchel, M. E., Hu, T. W., McDonnell, N. S., & Swisher, J. D. (1984). Cost-effectiveness analysis of an educational drug abuse prevention program. Journal of Drug Education, 14 (3), 271-292.

National Household Survey on Drug Abuse. (1996). Rockville, MD: U.S. Department of Health and Human Services.

O’Donnell, J., Hawkins, J. D., Catalano, R. F., Abbott, R. D., & Day, L. E. (1995). Preventing school failure, drug use, and delinquency among low-income children: Long-term intervention in elementary schools. American Journal of Orthopsychiatry, 65 (1), 87-100.

Pentz, M. A., Dwyer, J. H., MacKinnon, D. P., Flay, B. R., Hansen, W. B., Wang, E. Y., & Johnson, C. A. (1989). A multi-community trial for primary prevention of adolescent drug abuse: Effects on drug use prevalence. Journal of the American Medical Association, 261, 3259-3266.

Perrine, M., Peck, R., & Fell, J. (1988). Epidemiological perspectives on drug driving. In Surgeon General’s workshop on drug driving: Background papers. U.S. Department of Health and Human Services.

Plotnick, R. D. (1994). Applying benefit-cost analysis to drug use prevention programs. International Journal of the Addictions, 29 (3), 339-359.

Rice, D. P., Kelman, S., & Miller, L. S. (1991). Economic costs of drug abuse. In W. S. Cartwright & J. M. Kaple (Eds.), Economic costs, cost-effectiveness, financing, and community-based drug treatment. NIDA Research Monograph 113 (pp. 10-32). (DHHS Publication No. ADM 91-1823). Washington, DC: U.S. Government Printing Office.

Rice, D. P., Kelman, S., Miller, L. S., & Dunmeyer, S. (1990). The economic costs of alcohol and drug abuse and mental illness: 1985. (DHHS Publication No. ADM 90-1694). Washington, DC: U.S. Government Printing Office.

Rohrbach, L. A., Hodgson, C. S., Broder, B. I., Montgomery, S. B., Flay, B. R., Hansen, W. B., & Pentz, M. A. (1994). Parental participation in drug abuse prevention: Results from the Midwestern Prevention Project. Journal of Research on Adolescence, 4 (2), 295-317.

Russell, L. B., Siegel, J. E., Daniels, N., Gold, M. R., Luce, B. R., & Mandelblatt, J. S. (1996). Cost-effectiveness in health and medicine. In M. R. Gold, J. E. Siegel, L. B. Russell, & M. C. Weinstein (Eds.), Cost-effectiveness analysis as a guide to resource allocation in health: Roles and limitations. New York: Oxford University Press.

Scheffler, R. M., Feuchtbaum, L. B., & Phibbs, C. (1992). Prevention: The cost-effectiveness of the California Diabetes and Pregnancy Program. American Journal of Public Health, 82 (2), 168-175.

Scheffler, R. M., & Paringer, L. (1980). A review of economic evidence on prevention. Medical Care, 18 (5), 473-484.

Schweinhart, L., Barnes, H. V., & Weikart, D. P. (Eds.). (1993). Significant benefits: The High/Scope Perry Preschool Study through age 27. Ypsilanti, MI: High/Scope Educational Research Foundation.

Shi, L. (1993). A cost-benefit analysis of a California County’s Back Injury Prevention Program. Public Health Reports, 108 (2), 204-211.

Sindelar, J. L. (1991). Economic cost of illicit drug studies: Critique and research agenda. In W. S. Cartwright & J. M. Kaple (Eds.), Economic costs, cost-effectiveness, financing, and community-based drug treatment. NIDA Research Monograph 113 (pp. 33-45). (DHHS Publication No. ADM 91-1823). Washington, DC: U.S. Government Printing Office.

Spoth, R., Redmond, C., Haggerty, K., & Ward, T. (1995). A controlled parenting skills outcome study examining individual difference and attendance effects. Journal of Marriage and the Family, 57, 449-464.

Stein, J. A., Swisher, J. D., Hu, T. W., & McDonnell, N. (1984). Cost-effectiveness evaluation of a Channel One Program. Journal of Drug Education, 14 (3), 251-269.

Swint, J. M., & Nelson, W. B. (1977). Prospective evaluation of alcoholism rehabilitation efforts: The role of cost-benefit and cost-effectiveness analysis. Journal of Studies on Alcohol, 38 (7), 1386-1404.

Teutsch, S. M. (1992). A framework for assessing the effectiveness of disease and injury prevention. Morbidity and Mortality Weekly Report, 41 (RR-3), 5-13.

Torrance, G. W., Siegel, J. E., & Luce, B. R. (1996). Framing and designing the cost-effectiveness analysis. In M. R. Gold, J. E. Siegel, L. B. Russell, & M. C. Weinstein (Eds.), Cost-effectiveness in health and medicine (pp. 54-81). New York: Oxford University Press.

Tricker, R., & Davis, L. G. (1988). Implementing drug education in schools: An analysis of the costs and teacher perceptions. Journal of School Health, 58 (5), 181-185.

Weisbrod, B. (1983). A guide to benefit-cost analysis as seen through a controlled experiment in treating the mentally ill. Journal of Health Politics, Policy and Law, 7 (4), 808-845.

Winsdor, R. A., Lowe, J. B., Perkins, L. L., Smith-Yoder, D., Artz, L., Crawford, M., Amburgy, K., & Boyd, N. R. (1993). Health education for pregnant smokers: Its behavioral impact and cost benefit. American Journal of Public Health, 83 (2), 201-206.

Wylie, W. E. (1983). Cost-benefit analysis of a school health education program: One method. Journal of School Health, 53 (6), 371-373.

Yamada, T., Kendix, M., & Yamada, T. (1996). The impact of alcohol and marijuana use on high school graduation. Health Economics 5, 77-92.

ACKNOWLEDGMENTS

This literature review was written for the National Institute on Drug Abuse’s Resource Center for Health Services Research. In writing this review, the authors drew on their experience developing costs and benefits methods for drug abuse prevention research as part of the Alpha Prevention Project. The authors wish to gratefully acknowledge the guidance and support of their colleagues in the Alpha Prevention Project, especially Shirley Coletti, Randy Ratliff, and Joe Citro at Operation PAR, Inc.; Howard Hinesley, Steve Iachini, Dee Walker, and Deril Wood at Pinellas County Schools; Linda Miller and the staff of the ALPHA Program; Daphne Lampley, Nancy Helt, Mary Wright, and the staff of the Alpha Prevention Project; David Salkever and Marsha Lillie-Blanton at Johns Hopkins University; and Joe Takeda and Kellie Douglas at The University of Tennessee Children’s Mental Health Services Research Center. Special thanks go to Joe Takeda for his work on bibliographic retrieval, to Kellie Douglas for her work on manuscript preparation, and to Carol Schreter for scientific editing.