Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Executive Summary In an expansive view of the drug problem, drug abuse prevention re- search could be seen as a burgeoning domain, encompassing nearly every type of research with a bearing on individual health and social well-being, from the molecular to the global. However, the purview of this report is not nearly so expansive. Its purpose is threefold: . Assess the self-designated drug abuse prevention strategies that have been subjects of evaluation research, which are limited largely to a few domains of health-oriented interventions; · Consider the explicit theoretical basis and methodological adequacy of these evaluation findings and assess their applicability to diverse popula- tion segments; and Proffer minimum methodological standards for future evaluation projects. Within this scope, as defined by the sponsor of the study, the National Institute on Drug Abuse (NIDA), the committee has framed a limited set of conclusions concerning the direction of future research. The literature re- viewed in this report is devoted nearly entirely to studies of youth under age 20 and psychoactive drugs that are illegal for young people to purchase: the fully illicit drugs such as marijuana, heroin, and crack cocaine; the "prescription-only" drugs such as barbiturates and amphetamines; and the "adults-only" drugs, cigarettes and alcohol. The following summary responses to the specific points of NIDA's charge reflect our reading of this literature in the light shed by scientific principles, keeping in mind the pragmatic challenges of conducting research with hu 1
2 PREVENTING DRUG ABUSE: WHAT DO WE KNOW? man subjects in real social institutions on a topic bristling with emotional and political thorns. Review the current status of drug abuse prevention research; assess the theoretical basis for preventive interventions as derived from etiologic research. Research on drug abuse prevention is haunted by a double vision that emerges from epidemiologic studies. There seem to be two worlds of drug abuse. In one world, that of relatively low-intensity consumption (drug use) among individuals who can be found in schools and households, drug experience is self-reported more frequently by the wealthy than the less wealthy and by whites than Hispanics or blacks. In this world, there have been steady and cumulatively very marked declines in the prevalence of marijuana use since the late 1970s and of cocaine since the middle 1980s, and heroin use is so rare as to be barely measurable. In another world, that of emergency rooms, morgues, drug clinics, juvenile detention centers, jails, and prisons, in which indicators of intensive drug consumption (abuse and dependence) are collected: the poor predominate, blacks and Hispanics ap- pearing in numbers much higher than their household or school proportions; marijuana and heroin use are common (though less so in some areas than in the 1970s); and cocaine use increased explosively throughout the 1980s and simply leveled off at high levels in the l990s. The validity of the data that define each of these worlds, although subject to some degree of error and drift, is beyond knowledgeable dispute. Reconciling the two worlds in terms of theoretical understanding and em- pirical mechanisms, however, is a major research issue. To some degree these discrepancies may represent time lags, as tidal changes in the social acceptability and marketing of illicit drugs work their way through age- specific multiyear developmental pathways that lead from more or less common experimental use to a much smaller residual core of chronic drug depen- dence. But more of the discrepancy appears attributable to deep-seated divisions between the circumstances and social reinforcements of rich and poor, ethnic/linguistic majorities and minorities, and individuals predisposed toward or against strong attachments to drug-taking behaviors even before the opportunities to use specific drugs arise. A major finding of etiologic research is that the onset of drug taking follows relatively orderly sequences, which begin in early adolescence with the illicit use of alcohol and tobacco drugs widely and legally available to adults although prohibited to minors and end for some in a glut of drug consumption including the above and extending to cocaine and possibly heroin. For this reason, efforts to stop or at least delay to older ages the onset of use of these drugs, as well as efforts to act directly against mari
EXECUTIVE SUMMARY 3 juana and cocaine use, are suggestive paths for interventions. However, etiologic research also gives strong reasons to think that early onset can mean very different things for youths whose social supports are strong and relatively untroubled, than for those whose social environment is impover- ished or antagonistic and whose behavior includes a substantial repertoire of illegal and hazardous activities. As a result, the research suggests that prevention may need to proceed along distinct paths and that interventions may prove to have contradictory effects null for some, appreciable for others, even negative as well as positive directions of change in desired outcomes for different subpopula- tions. Etiologic studies further tell us that these populations are sorted and shaped in their knowledge, attitudes, and behavior by the people in whose presence (both personal and impersonal) they spend their lives. There is, in particular, a substantial deficit of information about how schools and com- munities two major youth-affecting institutions do this shaping and sort- ing, and how preventive interventions delivered person-to-person and through mass communications media interlock with the dynamic life of schools and communities. Strategic research initiatives are needed to improve our un- derstanding of the normative and economic aspects of communities and the normative and socioenvironmental character of schools and other institu- tions, as they affect drug-related and other health behavior, in order to prime the next generation of prevention strategies. Identify which drug abuse prevention strategies have been adequately evaluated and found to be effective, not effective, and countereffective. On balance, we conclude that no drug abuse prevention activities have been adequately evaluated and found to be reliably effective, in all cases, with all groups. One near-exception arises, in which a critical mass of findings of effectiveness are vitiated by methodological doubts and tem- pered by questions about the persistence and homogeneity of observed ef- fects: interventions in school settings from the 6th through 10th grades, focusing on behavioral training of skills to assertively counteract or resist (and, implicitly, to desist from exerting) explicit peer pressure toward use, lodged within a more general curriculum emphasizing self-efficacy, inter- personal social skills, and specific knowledge of health effects, followed up with booster sessions in a subsequent school year, and concomitant with continuing public health efforts on a community-wide basis, have in a no- table number of trials been effective in delaying the onset of cigarette smok- ing for a sizable fraction of students who would otherwise have begun smoking early in their adolescence. Although this seems a consistent enough finding to merit notice, there are important codicils. In controlled experimental studies begun long enough
4 PREVENTING DRUG ABUSE: WHAT DO WE KNOW? ago to yield follow-ups of 5 years or more, the deceleration in onset of cigarette smoking by students in the first year or more after exposure to the intervention does not necessarily yield lower smoking rates by the time students reach the upper classes of high school. Later interventions, using more technically refined approaches, may or may not prove to sustain these effects. However, even a delay in onset of smoking is noteworthy. Ciga- rette smokers who begin smoking later are likely to quit smoking sooner, and if smoking precedes onset of other drugs, later smoking means later onset of other drugs for which similar patterns (start later quit sooner) apply. More troubling about these studies are indications that the effects are not uniform; in exemplary, rigorously controlled evaluation, the students who had already begun smoking before receiving the curriculum became more likely to continue smoking afterward, even though the students who had not begun smoking were less likely to start afterward. The rates of attrition in these studies, particularly due to their reliance on school-based sampling, leave these subgroup results somewhat unsettled. Nevertheless, these negative findings point to countereffectiveness within the subpopula- tion described earlier as the second world of etiologic risk, and the positive results match findings elsewhere supporting effectiveness with the first- world population. Some prevention strategies have been evaluated sufficiently to con- clude that they are not widely effective. The will to believe on the part of implementers and program sponsors alike seems stronger than the evidence supports. This applies in particular to those school-based activities that do not at any point deal directly with the training of behavior between peers, but rather focus only on increasing knowledge about health effects, improv- ing interpersonal skills, or improving feelings of self-esteem. For drug abuse prevention strategies that have been found to be effective, assess how practical such strategies are for use in wide- scale applications and with other population groups (i.e., minori- ties). Because we cannot count any strategies as clearly and consistently ef- fective, the committee considers this point moot. Wide-scale programs must be conceived and executed as multiple strategies, each tailored to the specific population group it seeks to influence. Identify which prevention strategies have unknown effectiveness because of inadequate evaluation (i.e., insufficient numbers of replications). Although no inventory of prevention strategies has been taken that would identify the largest absorbers of funding, several school-based, skills-train
EXECUTIVE SUMMARY s ing-inclusive curricula (in particular, DARE and the Here's Looking at You series) are so widely employed that we must recommend the completion of additional rigorous evaluations by independent researchers or these strate- gies should cease being used. Were resources unlimited, we could well call for generous ladles of funding to support more and better evaluations of a substantial range of other intervention approaches. However, we find that, on the whole, an insufficient number of replications is not the main obstacle to identifying effective strategies. There seems much more warrant for formative, rela- tively smaller-scale studies using trials and other methodologies of pre- vention strategies based on theoretical principles such as risk-factor reduc- tion and developmental shaping of behavior. At the same time, broad-scale community strategies and conditions of living need research attention, for it is within the broader community context that any school-based or other strategy must operate and trace its effects. Review methodological issues regarding drug abuse prevention strategies: identify major design and methodology problems in evaluating ex- isting prevention strategies and possible approaches for correcting such problems in current and future prevention research. A clear majority of the research published as evaluations of the effec- tiveness of preventive interventions is methodologically weak. To a certain degree, this is an unsurprising result of imbalance between the large volume of prevention-oriented activities and the modest volume of support for their evaluation. A catalog of these weaknesses would be tiresome and perhaps misleading-there are a respectable number of sound studies. However, the most common, fundamental problems, which afflict even some of the most widely cited research, are as follows: Cursory description and documentation of the intervention methods the evaluation designs, the outcome measures used, and the characteristics of treatment and control populations, in terms of both personal characteris- tics and social circumstances. . Partial or missing measurement of instrumental processes and inter- mediate and final effects, including individual exposure to materials, mes- sages, or training; retention of knowledge; acceptance of affective or attitu- dinal impressions; changes in assertive or other behavior; changes in levels of drug consumption; changes in drug-related sequelae. · No attention to concurrent prevention activities in the experimental or control locales. Inadequate follow-up, insufficient time frames, and response rates that are too low and subject to serious biases. .
6 PREVENTING DRUG ABUSE: WHAT DO WE KNOW? The correction of such issues, insofar as that is within the power of NIDA, is not a matter of applying rigid formulae. It requires a patient commitment to attracting quality researchers to the field; applying require- ments such as discussed below to NIDA publications and encouraging other research sponsors, collaborators (such as school administrators), reviewers, and publication editors to attend to them; developing and supporting appro- priate research training; and attending to socioenvironmental aspects and data quality control elements of proposed research. Identify minimum requirements for assessing the effectiveness of prevention strategies. Evaluations of effectiveness must clearly specify and describe (directly or by reference to readily available supplementary sources or previous pub- lications) each of the following elements and make provision for quantita- tive measurement of each of them: · The components of the intervention strategy. Optimal and achieved levels of implementation of each component, from the perspective of both source and recipient. . . The prescribed and actual qualifications and training given to those implementing the strategy. · The levels and types of community and organizational support for and opposition to the intervention. · The character and extent of concurrent prevention activities in the research locale that affect the control and treatment subjects of the evalua- tion. . The specific cognitive, affective, and behavioral measures used to assess outcomes. · The characteristics of the treatment and the control populations un- der study (when present, whether randomly assigned or otherwise selected), including age, place of birth, sex, racial and ethnic identity, household structure and stability, household socioeconomic measures (household income and education level if available; other indicators such as residential density, vehicles owned, household furnishings), academic grades, block-level geo- graphic information, drug experience, and previous exposure to or involve- ment in prevention activities. Follow-up measures must be made at least 1 year after the initial inter- vention is completed, preferably at longer intervals when the expected rates of target behaviors to be affected are low or the sample size is small. In- terim exposure to prevention curricula or other elements must be assessed. Participant follow-up rates of 90 percent or greater are needed to mea- sure accurately the effects on relatively uncommon outcomes such as regu
EXECUTIVE SUMMARY 7 tar cocaine use. In most cases this means that evaluation research designs must make provisions to retain identifiers of individual participants and locate them if necessary in environments discontinuous with the original site. An evaluation with a gross follow-up rate below 75 percent is of dubious validity to assess effects even on relatively common behaviors. Nonresponse analyses must be performed and reported. A final note, which is not visible in the research under review and is not strictly necessary for effectiveness determination. We believe that the practical value of any evaluation is substantially improved if its performers take the trouble to give a careful accounting or best possible estimate of the unit costs of implementing the intervention, separately from the costs of the research components.