2
Principles of Harm Reduction

A broadly shared goal of public policy toward cigarettes and other tobacco products is to reduce their health burden (IOM 1994, 1998). That health burden is minimized if no individual begins smoking and those who are currently smoking quit promptly (U.S. DHHS, 1988). However, quitting is difficult for most smokers and many adolescents will experiment with smoking; experimentation predictably leads a substantial fraction to become regular smokers (U.S. DHHS, 1994). Thus in addition to interventions aimed at prevention and at promoting immediate quitting, it is appropriate to consider interventions that aim to reduce the harm that the remaining population of smokers cause themselves and others by continued smoking. This is the underlying concept of harm reduction or harm minimization.

The term “harm reduction” has a variety of applications. It can refer to a policy or strategy (a set of laws and programs) or to specific interventions (e.g., an individual product innovation or dissemination effort). A harm reduction policy or intervention (a) explicitly assumes continuation of the undesired behavior as a possibility and (b) aims to lower the total adverse consequences, including those arising from continuation. In this use, the term describes an assumption and a goal rather than a result. It can also be used as a criterion for evaluating results; an intervention or policy is harm-reducing if it does in fact reduce the total adverse consequences. Finally, harm reduction can also be viewed as a framework, a way of thinking about dealing with a harmful behavior, since it requires analysis of a broader set of outcome measures than would otherwise be



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction 2 Principles of Harm Reduction A broadly shared goal of public policy toward cigarettes and other tobacco products is to reduce their health burden (IOM 1994, 1998). That health burden is minimized if no individual begins smoking and those who are currently smoking quit promptly (U.S. DHHS, 1988). However, quitting is difficult for most smokers and many adolescents will experiment with smoking; experimentation predictably leads a substantial fraction to become regular smokers (U.S. DHHS, 1994). Thus in addition to interventions aimed at prevention and at promoting immediate quitting, it is appropriate to consider interventions that aim to reduce the harm that the remaining population of smokers cause themselves and others by continued smoking. This is the underlying concept of harm reduction or harm minimization. The term “harm reduction” has a variety of applications. It can refer to a policy or strategy (a set of laws and programs) or to specific interventions (e.g., an individual product innovation or dissemination effort). A harm reduction policy or intervention (a) explicitly assumes continuation of the undesired behavior as a possibility and (b) aims to lower the total adverse consequences, including those arising from continuation. In this use, the term describes an assumption and a goal rather than a result. It can also be used as a criterion for evaluating results; an intervention or policy is harm-reducing if it does in fact reduce the total adverse consequences. Finally, harm reduction can also be viewed as a framework, a way of thinking about dealing with a harmful behavior, since it requires analysis of a broader set of outcome measures than would otherwise be

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction considered. One cannot usually determine in advance, on theoretical grounds, whether a particular policy or intervention is harm-reducing. For example, it may turn out that a policy which aims to minimize prevalence (i.e., addresses only abstinence) reduces total harm as compared to any other policy. But the framework allows consideration of alternatives to reduction in the number of users as a complement to abstinence. The concept of harm reduction has application in a number of policy areas apart from tobacco, including automobile safety, sex education for children, alcohol control and policy toward illicit drugs. In some instances the harm reduction considerations are only implicit, providing an ex post rationalization of decisions already made (e.g., automobile safety, MacCoun, 1998). In others (e.g., needle exchange programs) it is a very prominent element of policy discussions. While none of the harm reduction interventions in these other policy areas are exactly analogous to those in the tobacco field, they will be used to illustrate the potential strengths and weaknesses of this approach for tobacco. The next section elaborates the basic framework of harm reduction and briefly relates it to risk assessment. The third section also shows some of harm reduction’s applications in related areas and compares these applications to some potential smoking interventions, though a much more extended discussion of those interventions is provided later in the report. The chapter concludes with some observations on the difficulties of applying the harm reduction framework, in particular the problems of developing measures to establish whether harm has in fact been reduced, and the need to give greater weight to mistaken acceptance of a product as harm-reducing than to mistakenly rejecting a harm-reducing product. CONCEPTUAL FRAMEWORK FOR HARM REDUCTION Basic Concepts Harm reduction accepts that interventions focused on reducing the harmfulness of a substance or behavior, even if they increase the extent of substance use or the frequency of the targeted behavior, may be able to reduce the aggregate of adverse consequences for society, including both users and nonusers. For example, referring to the alcohol field, a prominent group of researchers stated: “Unlike ‘abstinence-only’ or ‘zero-tolerance’ approaches, the harm reduction model supports any behavior change, from moderation to abstinence, that reduces the harm of problems due to alcohol” (Marlatt et al., 1993). For illicit drugs “the central defining characteristic of harm reduction is that it focuses on the reduction of harm as its primary goal rather than reduction of use per se, secondly that strategies are included to reduce the harms for those who continue to

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction use drugs, and thirdly that strategies are included which demonstrate that, on the balance of probabilities, it is likely to result in a net reduction in drug-related harm” (Lenton and Single, 1998). Frequently definitions assess alternatives not in aggregate health measures but in terms of the harms associated with a single act or product or with the individual user; for example, “the term harm reduction originally referred to only those policies and programmes which attempted to reduce the risk of harm among people who continued to use drugs” (Lenton and Single, 1998, citing Single and Rohl, 1997). The harms consist of all the adverse consequences borne by members of society. These include increased morbidity and mortality (among both users and nonusers) from all sources; addiction itself; expenditures on regulation or enforcement, since these are costs borne by taxpayers; the increased intrusiveness of the state; and crime that might be generated by regulation or enforcement or by the behavior itself. These adverse consequences are borne by many different groups: users themselves; intimates of the user, particularly children and spouses; nonusers directly (e.g., through crime, in the case of illicit drugs, or traffic accidents in the case of alcohol); and nonusers indirectly or society generally (e.g., through taxation). The value society gives in considering interventions to the interests of these groups may vary (MacCoun et al., 1996); typically a greater consideration is given to the welfare of children or of neonates, since they are the most vulnerable victims, with very limited capacity to undertake actions in their own interests. Analytics Total harm can be expressed as a function of the number of individuals engaged in the behavior and the damage each causes. In turn, the damage caused by an individual is a function of intensity of use (or frequency of behavior) and of the harmfulness of each episode of use or behavior. MacCoun and Reuter (2001) suggest that total harm can then be expressed as Total Harm=Harmfulness (per use)×Intensity (per user)×Prevalence (of use) It can be reduced through declines in any of the three components individually, including intensity. However the three components may not be independently determined. In particular, prevalence may be affected by harmfulness through three distinct, though related, paths; initiation, nondesistance, and relapse. The lower harmfulness may reduce perceived harmfulness and encourage someone to begin using drugs, to drive a car too fast, or engage in sex at too early an age. Perceived dangers may be influenced both by the actual

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction dangers and by information about those dangers. Where the substance or behavior is addictive or habit-forming, this may generate a long-term increase in the number of users. In addition, lower perceived harmfulness may reduce incentives to quit or desist. Relapse may also be encouraged by the perception of less dangerous means of continuing the desired behavior. Lower perceived harmfulness may also increase intensity of use. Changing the riskiness of an act is known to alter the behavior of the population. Generally the change is in the form of compensation, i.e., higher risk will reduce the prevalence of the behavior while lower risk will increase that prevalence. Engineers tend to overestimate the benefits from safety devices, since they ignore that behavioral adaptation. On auto safety, Evans (1991) has noted: “If the safety change affects vehicle performance, it is likely to be used to increase mobility. Thus, improved braking or handling characteristics are likely to lead to increased speeds, closer following, and faster cornering. Safety may also increase, but by less than if there had been no behavioral response.” In its most extreme form, this kind of risk compensation has been labeled risk homeostasis—a term that implies implicit or explicit efforts to maintain a constant level of risk (Wilde, 1982). These changes are logical possibilities. How substantial they are is an empirical matter. So is the extent of generalizability across product domains and populations. The introduction of a safer automobile may, for example, have negligible effect on the driving behavior of older drivers but sharply increase speeding by younger drivers, while condom availability may increase sexual behavior of older adolescents more than it affects that of younger adolescents. The psychological mechanisms generating alcohol and cigarette dependence may be different enough that harm reduction interventions in general are more effective in one field than in the other. This lack of generalizability raises a question as to the relevance of examples from other fields to tobacco interventions. But tobacco harm reduction involves a large variety of potential interventions. They differ in some important dimensions, just as do automobile seat belts and needle exchange programs. The examples can help identify the dimensions that influence the outcomes of harm reduction interventions. APPLICATIONS The harm reduction framework can be applied not only in a number of different policy areas but to a variety of forms of interventions: Lowering the inherent harmfulness of a broad class of products

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction (automobile safety regulation); this is the approach used for tobacco harm reduction products that rely on removal of some toxicants from tobacco. Shift to less toxic mode of ingestion. Needle exchange programs attempt to reduce the harmfulness of the act of injecting drugs, without requiring abstinence. That is the approach embodied by products such as Eclipse, with heated tobacco or tobacco-like materials providing nicotine by a similar mechanism that allows continuation of the act of smoking but attempts to make it less harmful. Behavioral change therapies (controlled drinking); many tobacco harm reduction strategies will require behavior change as a complement to product innovation. Adding a less harmful but dependency-creating product to the available mix of dangerous products (methadone for heroin addicts); this is the rationale for nicotine replacement therapy for long-term use. This section provides a brief description of the nature of each of the non-tobacco interventions listed and how they have fared in the harm reduction framework. It also describes the extent to which they have had the effects predicted for them when introduced. The examples presented are, by nature, imperfect analogies of tobacco harm reduction but are offered to highlight the positive and negative implications of harm reduction interventions. Automobile Safety Regulation Automobiles are a source of numerous injuries and fatalities; in 1998 there were 41,200 deaths in the United States associated with automobiles (National Safety Council, 1999). Some of these, but not all, are a consequence of unsafe operation of vehicles, in particular driving while intoxicated or driving at high speed. The National Highway Traffic Safety Administration (NHTSA) estimates that speeding was a contributing factor in 30% of such deaths (NHTSA, 1999a). It was also estimated that in 1994, 16,600 traffic fatalities were alcohol-related (CDC, 1995). A series of product innovations, including seat belts, anti-lock braking systems, and air bags have led to large improvements in the crash-worthiness of vehicles. Lap/shoulder safety belts, when used, reduce the risk of moderate to fatal injury to front seat passenger occupants by approximately half (NHTSA, 1999b). Federal law now requires their installation in new vehicles and all states mandate that they be used. From the earliest days of these innovations, there has been a research interest in behavioral adjustments that might reduce the effectiveness of these innovations. Given that cars are safer, drivers may be more inclined to exceed the legal speed limit as well as exercise reduced care with respect

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction to driving while alcohol impaired. This is analogous to increased intensity of smoking when given access to low tar and nicotine cigarettes. Peltzmann (1975) found that changes in driver behavior more than offset product improvement. However, since then a substantial literature has refuted that finding. The research generally concludes that there is in fact more speeding but that the net result is a reduced burden of automobile injury and mortality. For example, Chirinko and Harper (1993) found that the introduction of air bags reduced automobile fatalities by between 13.8 and 26.1%. They also observed a shift (as did Peltzmann) in the composition of those fatalities; non-occupant fatalities (i.e., deaths of pedestrians, cyclists, etc.) increased while occupant fatalities decreased. This latter finding is consistent with the hypothesis that the technology encourages less safe driving. Other engineering improvements on roads (stronger guard rails, brighter lights) have also reduced the frequency of accidents or severity of injuries associated with unsafe driving (Ross, 1992). Evans (1991) notes that: “When safety measures are largely invisible to the user, there is no evidence of any measurable human behavioral feedback. Likewise, when measures affect only the outcome of crashes, rather than their probability, no user responses have been measured.” These are instances of pure harm reduction interventions; no behavioral response diminishes their design effect. Seat belts can be disabled; like many harm reduction interventions they require compliance for their effect. One factor explaining the relative modest impact of mandatory seat belt laws on traffic fatalities lies in substantial rates of non-compliance (Dee, 1998). Moreover the population is not homogeneous: “unsafe drivers may be the least likely to adjust their belt use after the introduction of the law” (Dee, 1998). Thus the safety effects of automobile innovations are less than expected due to both increased speeding and selective noncompliance regarding seat belts. Though there is strong and increasing social disapproval of unsafe driving, as expressed in congressional passage of legislation in 2000, urging states to reduce the maximum allowable Blood Alcohol Content (BAC) to 0.08%, there is no public discussion that these mandated product changes might encourage faster or less safe driving or redistribute the damage toward pedestrians and other innocent parties. This lack of debate may reflect, inter alia, the compelling nature of the intervention, reducing the risks associated with being the driver of a car, an almost universal experience of American adults. What is immediately discernible is the reduction for the driver, not the potential increase for other parties. Moreover, the targeted behavior (unsafe driving) is so common that, even though there is support for increasingly stringent laws and enforcement,

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction there is an acceptance of its inevitability; there are no calls for a “speeding free society.” Speeding, though not driving while intoxicated, also confers a benefit—namely, reduced travel time. These benefits of reduced mortality and injury have been obtained at considerable cost. The estimated cost per life-year saved varies widely for different devices and mandates; for example, the figure for driver airbag and manual lap belts (as compared to manual lap belts alone) is $6,700 while for the same devices for passengers and drivers, the figure is $62,000 (Tengs et al, 1995). Some interventions are rated as essentially costless; for example, driver automatic (vs. manual) seat belts. These require investments only of government authority, not financial resources. Automobile safety also illustrates the potential conflict among social goals that may be ignored even in a harm reduction framework. To reduce gasoline consumption and air pollution, the corporate average fuel economy standard (CAFE) has led to lighter cars; these cars are correspondingly less safe (Crandall and Graham, 1989). The trade-off between the two goals (pollution abatement and reducing injury and death from automobile accidents) has not been evaluated and is the subject of little public discussion. Indeed, the National Highway Traffic Safety Administration has been accused of obfuscating precisely this issue (Kazman, 1991). Harm reduction, despite its ostensible breadth, does not necessarily cover all potential adverse consequences. Automobile safety represents an instance of a successful harm reduction intervention. There is indeed an increase in harmful behavior but not so much as to overcome the reduction in adverse consequences of that behavior. Teen Sexual Behavior Births to teenagers have been identified as a major societal health and behavioral problem (Ventura et al., 1997). For example, teenage mothers are less likely to complete their own education or provide adequate parental supervision and more likely to give birth to a low-birthweight infant. Over one third of teenage pregnancies end in abortions (Henshaw, 1999). Unwanted children are particularly at risk of neglect from teenage mothers (Federal Interagency Forum on Child and Family Statistics, 1997). There is a societal interest in reducing the number of unwanted infants born to teenage females. In addition, teen sexual activity, which frequently occurs outside of monogamous relationships, also facilitates the spread of sexually transmitted diseases (STDs), including AIDS. The severity of problems related to premature sexual intercourse is a function of the prevalence of the acts and their average safety—mostly the probability of pregnancy and of STDs. Until recently many schools chose

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction to emphasize the value of abstinence before a certain age; not only does that represent community values but abstinence, if achieved, eliminates the risky acts (Kirby, 1992). However, the rise in sexual activity among adolescents and the emergence of the AIDS epidemic have led to a concern that “abstinence only” messages may be ineffective and that the adverse consequences of unprotected sex may be greater than previously estimated. There has been exploration of the effects of harm reduction interventions which accept the high probability of sexual activity at an inappropriately young age and aim to reduce the probability of pregnancy or disease (Kirby, 1997). Such interventions can take a number of forms. One is curricula aimed at teaching adolescents about responsible behavior, if they choose to have sexual relations. Such a curriculum may reduce the adverse consequences of early sexual activity. Opponents of these programs believe that they may encourage early sexual activity, in itself socially undesirable. However, a recent review of evaluations of a number of curricula concluded that “none of the 11 studies that examined the impact of programs on the frequency of intercourse found a significant increase” (Kirby and Coyle, 1997). Another possible method of reducing harm is to facilitate access to adequate birth control technologies to adolescents, in particular to school children. Kirby and Brown (1996) estimate that by the mid-1990s nearly 400 schools made condoms available to students. Four recent studies found little evidence that the programs raised the percentage of high school students who engage in sexual intercourse, either by increasing teen awareness of how frequent such activity is among their peers or by reducing its perceived risks (Kirby and Coyle, 1997). On the other hand, evaluations of these programs also provide mixed support as to an effect on utilization of condoms, or any other birth control device. For example, a study of nine Philadelphia schools which provided reproductive health information, condoms, and general health referrals found that these schools showed no significant increases or decreases in condom use over time as compared to schools which did not install these programs (Furstenberg et al., 1997). A study of Seattle schools which made condoms available in vending machines or in baskets at school health clinics concluded that students use a relatively large number of condoms distributed in this fashion, but this “did not lead to increases in either sexual activity or condom use” (Kirby et al. 1999). The condoms distributed through the schools may have substituted for others obtained through other channels. There was one discrepant study: a condom distribution program that was part of a comprehensive AIDS prevention program increased rates of condom use in a sample of Chicago schools compared to a set of matched controls (Guttmacher et al., 1997).

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction These interventions illustrate the effects of increasing the availability of a potentially harm-reducing product to a specific high-risk population. It appears that this one has slight impact on behavior, on either the frequency of the acts or their average safety. In the absence of a strong scientific base, the debate is largely in terms of values and impressions, a common characteristic of social policy fields in which harm reduction has been applied. Making condoms available to school students may “send the wrong signal.” It involves the state in apparently facilitating acts of which society disapproves. Since the evidence to date is that these interventions have at most a modest effect on the frequency and damage of the targeted behavior, the harm reduction framework has not been explicit. Alcohol Alcohol policy raises many harm reduction issues, reflecting the mixed social message with respect to alcohol’s health consequences. Light drinking is a socially acceptable behavior, with apparently health promoting consequences. Heavy drinking, particularly chronic heavy drinking, is the source of a huge burden of morbidity and mortality, and is acknowledged as dangerous both to the drinker and others. Harm reduction enters alcohol control in a number of ways. Alcohol consumption is characterized, even for most light drinkers, by episodes of excessive drinking. Though each light drinker has only a small to moderate probability of an alcohol-related automobile accident or other kind of injury, their numbers are large enough that, as a group, light drinkers account almost half the damage associated with alcohol consumption (Kreitman, 1986). Duncan (1997) found that driving while intoxicated (DWI) rates across states were associated with binge drinking but not with chronic heavy drinking. As a result, a central debate is whether alcohol policy should focus on heavy drinkers as a group, on the total amount of alcohol, consumed or primarily on intoxication as a behavior. For example, some programs in the last group emphasize how the potential harm of drinking a given amount can be reduced by consuming it over longer periods of time or eating food during the drinking session. That implies an acceptance of heavy drinking, itself an unhealthy behavior and one that is a risk factor for numerous diseases. In contrast, high alcohol taxes reduce aggregate consumption, including that which is nonharmful; these taxes can be seen as “punish[ing] the many for the sins of the few” (Stockwell et al, 1996). If all drinking is seen to generate some probability of adverse effects, as the total consumption model suggests, then measures that reduce overall drinking are likely to be harm-reducing. Single (1997) suggests that “[I]ncreased attention is likely to be given to prevention measures which

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction focus on preventing problems associated with drinking rather than restricting access to alcohol.” This is only partly rooted in public health; it also represents the political realities of decreasing public support for restricting alcohol availability and the dissemination of data showing that moderate levels of alcohol consumption appear to reduce all-cause mortality (Edwards, 1995; NIAAA, 2000). Harm reduction shows up in other aspects of alcohol policy; examples include “measures to reduce nonbeverage alcohol by ‘Skid Row’ inebriates, measures to reduce intake of alcohol by drinkers (e.g., promotion of low-alcohol beverages, server training programs) and measures to reduce the consequences of intoxication” (Single, 1997). For example, impoverished, single men who have chronic alcohol problems are at risk of drinking various liquids that contain alcohol but are not fit for human consumption (e.g., methylated spirits). Since these men are unable (both for financial reasons and because of poor self-control) to maintain stocks of alcohol, they are likely to consume these more dangerous substances if they cannot readily obtain alcoholic beverages, as happens early in the morning when liquor stores are closed. One method of alleviating this problem is to expand opening hours for stores operating in Bowery-like areas. Early opening may reduce the extent of drinking of nonbeverage alcohol and thus alcohol-related morbidity and mortality. However, it also signals society’s willingness to facilitate drinking by individuals with serious drinking problems who will remain untreated and some of whose problems may be exacerbated or ameliorated by allowing easier access to alcohol. Some harm reduction interventions in the alcohol field are indeed aimed directly at harms and seem unlikely to induce behavioral responses that would ameliorate their effects. For example, intoxication leads to numerous violent fights in pubs and bars. In Scotland, one intervention involved serving drinks in glass which crystallized rather than shattered when it broke, thus reducing the damage caused by such fights (Plant et al., 1994). That seems highly unlikely to increase the extent of fighting (the proximate source of harm) or heavy drinking (the more distal source of the harm); indeed, if the purpose of fighting is to cause the maximum injury to others, it may actually reduce the prevalence of fighting. Harm reduction also operates at the clinical level. A long-standing belief among treatment specialists is that any message for individuals with drinking problems other than abstinence imposes unacceptable risk of relapse to dangerous drinking behavior. Over the last quarter century, however, a number of studies have found that controlled drinking may be a better goal for at least some problem drinkers (e.g., Miller and Caddy, 1977; Sobell and Sobell, 1978), though the results have not been consistent (Foy et al., 1984; Vaillant, 1996). Its appropriateness is in part a function of

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction targeting. Can one identify subpopulations of problem drinkers with higher probability of benefiting from goals other than abstinence? Illicit Drugs Harm reduction has been most explicitly discussed in terms of policy toward illicit drugs. A distinctive aspect of drug policy is that the policies themselves have direct adverse consequences. Toughly enforced prohibitions aimed at use reduction can lead to drug overdose deaths due to drugs of unknown purity and poor quality. That does not imply that the prohibitions should be relaxed. These current policies may in fact be harm-minimizing, since the high prices and limited accessibility reduce the use and volume of addiction- or intoxication-related harms. This presents an empirical issue that remains unexamined (Caulkins and Reuter, 1997). However the very prominence of the harms directly related to prohibition, such as the violence and disorder around drug markets has generated an interest in the possibility that society would benefit from less punitive policies, even if they increase prevalence of drug use. Two important interventions have been the subject of harm reduction debates: needle exchange programs (NEPs) to reduce the spread of AIDS and the provision of methadone to ameliorate the adverse consequences of opiate dependence. The spread of HIV among intravenous drug users (IVDUs) and their sex partners is primarily a function of needle sharing, not of the drugs consumed. NEPs aim to reduce harm caused by IV drugs by reducing the risky practice. Opponents of NEPs argue that these programs facilitate a dangerous and illegal behavior: IV drug use. The proponents of needle exchange programs argue that whatever its symbolism, both public health and considerations of humane treatment of drug addicts require NEP. There is a base of research demonstrating the positive public health effects. As summarized by a National Research Council panel, “NEPs increase the availability of sterile injection equipment. For the participants in a (NEP)…this amounts to a reduction in an important risk factor for HIV transmission…. There is no credible evidence to date that drug use is increased among participants as a result of programs that provide legal access to sterile equipment. The available scientific literature provides evidence based on self-reports that needle exchange programs do not increase the frequency of injection among program participants and do not increase the number of new initiates to injection drug use” (Normand et al., 1995). Several recent policy reports have upheld those conclusions (IOM, 2000; U.S. DHHS, 1998). There is also strong popular support as revealed in survey research (Henry J.Kaiser Family Foundation, 1996). Congress has been unpersuaded by either the research or the

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction moral arguments. The belief that it will “send the wrong signal” and thus increase drug use by undermining the abstinence message, both to current and potential users, does not seem to be responsive to empirical findings. Perhaps more relevant to the current concerns about harm reduction in the tobacco field is the introduction of methadone as a maintenance drug for heroin addicts. Methadone is a long-acting agonist that reduces craving for other opiates; see Rettig and Yarmolinsky (1995) for a review. Provision of low-cost methadone has enabled hundreds of thousands of opiate addicts in a dozen countries (North America, Western Europe, and Australia) to lead substantially better lives; they are able to avoid the humiliation of searching for an expensive prohibited drug, achieve modest levels of workplace functioning, and mitigate major threats to their own health and the health of others. It has helped limit the spread of HIV (Longshore et al., 1993). While it is an abusable drug, addicts on methadone have a much lower mortality rate than untreated addicts (Ball and Ross, 1991). Methadone dependence seems to be at least as difficult to end as heroin dependence. Patients who discontinue methadone use relapse to opioid dependence at a high rate (Ward et al., 1992). The perception of heightened dependence is common amongst patients. Some addicts use methadone when their heroin use has become particularly problematic, with the expectation of returning to heroin use when they are past this particular crisis. Methadone was a major ideological battlefront in the 1970s and 1980s (Rosenbaum, 1997; White House Conference for a Drug Free America, 1988). It has been noted that “the controversies over methadone treatment stem almost entirely from philosophical differences—objections to the substitution of one drug for another—and not from doubts about the pharmacological safety and efficacy of methadone” (Newman and Peyser, 1991). Methadone is a dependency-creating opiate, as is heroin. Methadone dependence is more acceptable than heroin dependence because it improves the user’s function as a member of society. That the methadone patient gets less pleasure from the substitute drug is probably not critical in itself, though the lessened intensity of pleasure may allow for greater engagement with others and less self-centeredness and thus helps generate popular acceptance. For some policy makers, however, the reduction in the burden of disease and other social dysfunction is not enough to justify government funding and provision of an addictive drug. In this case, in contrast to the battle over NEP, the proponents have prevailed. Methadone maintenance, though inadequately funded and poorly delivered, is the mainstay of the U.S. treatment for heroin addicts (Rettig and Yarmolinsky, 1995).

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction There is no evidence that methadone has increased initiation into opiate addiction where it has been made available. However, it is difficult to develop a powerful design for testing the hypothesis, particularly given the paucity of local indicators of heroin use. There has been no association between the number of persons in methadone programs and the number of reported new users in the National Household Survey on Drug Abuse, but that is at best a weak test. Assessing whether methadone prolongs a career of opiate dependence is difficult because of the very different characteristics of the methadone and heroin initiate populations. Even among addicts, many desist for long periods without treatment; for example, Anglin et al. (1986) found that 56.4% of a sample of 406 heroin addicts were able to desist for three years or more without formal assistance. However, those who initiate methadone use for addiction treatment are those who were unable to quit heroin; hence the difficulty of comparison. Summary of Applications Harm reduction has been controversial wherever it has been applied explicitly; moreover it often has a weak base in terms of assessed outcomes. Automobile safety is one instance for which there is good evidence of both compensatory behavior and net harm reduction. It is also the instance in which harm reduction issues have been least clearly articulated in public debate, though widely discussed in the traffic safety community. Interventions to reduce the danger of adolescent sexual behavior have shown no adverse effect, in terms of increased sexual activity, but also little indication that the interventions have reduced the average harmfulness of the acts. For alcohol, harm reduction is gathering momentum but with only a modest scientific basis at either the population or clinical level. For heroin, research on methadone and needle exchange programs provide evidence that they do reduce total harms resulting from use by currently dependent users; there is a weaker evidentiary base for concluding that initiation is unaffected. COMPARING OTHER HARM REDUCTION INTERVENTIONS TO THOSE FOR SMOKING These examples are offered for the insights they may provide as to the likely consequences of tobacco-related potential reduced-exposure products (PREPs). However, harm reduction can work through a number of mechanisms and has consequences in a number of dimensions. Assessing the relevance of these harm reduction interventions to PREPs requires consideration of those mechanisms and consequences.

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction Table 2–1 compares three types of tobacco-related PREPs with five other harm reduction interventions discussed above. It aims to show in what ways the various PREPs are similar to or different from the other interventions. It is intended to be illustrative rather than conclusionary; a number of entries are conjectural. The comparisons are in terms of: Theory of how the product/policy might reduce harm (Presumed Mechanism). Each intervention posits a specific mechanism as to how it might diminish adverse consequences at the individual level. For example, modified tobacco attempts to present a less toxic version of a familiar product, in contrast to nicotine replacement therapy (NRT), which involve a substitute product with a very different mode of consumption and action. Methadone is similar to NRT in that respect, while NEPs are not. This is not a judgment of actual success but simply of the theoretical basis for believing that it is possible the intervention reduces adverse consequences. Effect on prevalence of the undesired behavior. These are crude summaries of the empirical literature; where the result is particularly conjectural, this is indicated by a question mark. For modified tobacco products the entry reflects the almost certain effect of allowing their availability, with claims. The undesired behavior in the case of PREPs is smoking; for others it is fast driving, underage sex, injecting of powerful illegal opiates, and drinking by problem drinkers. This is an estimate of the effect on the number of undesired acts, not of the total harms themselves. Effect on intensity of use. As for prevalence, these are crude summaries of complex empirical literatures. Harms are a function of both prevalence and intensity of use. Low tar and filter cigarettes not only led more individuals to smoke but also on average led them to higher daily consumption. Early opening of bars on the Bowery may lead to higher alcohol consumption by chronic alcoholics, while needle exchange programs generally either reduce injecting frequency or leave it unchanged. Effects on others. Again, these are summaries of the empirical literature. In the case of mandated seat belts, the increase in speeding may have negative consequences for pedestrians, while reducing mortality of vehicle passengers or even total traffic-related mortality. Nicotine replacement products, even if they lead to more nicotine consumption, may reduce ETS sufficiently to lower harms to others. Whether the intervention conveys symbolic approval of the undesired behavior. Allowing cigarette manufacturers to market cigarettes with the claim of lower carcinogens requires the government to approve the act of smoking cigarettes, even if accompanied by warning signs, just as do condom programs for kids (underage sex) and needle exchange for addicts (injecting drug use). Other interventions have no such effect. NRT meets a physiological need through such different mechanisms that they

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction TABLE 2–1 Characteristics of Eight Interventions With Harm Reduction Rationales Product Presumed Mechanism Effect on Prevalence Harms to Others Symbolic Approval? Effect on Intensity of Use Potential Threats to Reducing Harm Light & filter cigarettes Less dangerous product Increased cigarette consumption Increased No (govt. action not required) Increase Adaptive behavior negates technology/raises prevalence Modified tobacco Less dangerous product Increased cigarette consumption Increased Yes (if claims allowed) Increase Exposure reductions not realized/ prevalence rises too much Nicotine replacement therapy Substitute product Decreased smoking? Reduced No Reduce Prolongs smoking careers/ incomplete compliance Oral methadone Substitute product Reduced heroin/raise opiate use Reduced No Lower heroin/ possibly higher opiate Prolongs opiate use Needle exchange programs Reducing danger of act None Reduced Yes Reduce/ unchanged Incomplete compliance Early opening of bars on the Bowery Shifting consumption from dangerous forms None Unchanged Yes Increase Behavioral model incorrect Adolescent condom distribution Reducing danger of act Unclear Unclear Yes Unclear Increased sexual activity among kids Mandated seat belts Safer product Increase Unclear No Increase Faster driving/selective compliance

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction constitute no approval of smoking. Oral methadone also represents no endorsement of injecting powerful opiates, it may prolong opiate use but in a form that permits social functioning. This dimension may be important for establishing popular and professional acceptability. An assessment of why a harm reduction intervention or PREP might fail to reduce total harm. There are many paths to failure. For example, even if the mechanism of modified tobacco products is correct and they are less toxic than conventional products, the population changes may be so great as to lead not to harm reduction but to greater total adverse consequences. In the case of NEPs, it may be that compliance is so incomplete, or that the HIV epidemic is so far advanced, that needle exchange programs fail to have any detectable effect. The Skid Row intervention may underestimate the effect of early opening on other heavy drinkers. The table indicates that no two interventions are identical in all dimensions. If one accepts the utility of distinguishing between making an existing product safer and offering a very distinctive substitute, only mandated seat belts parallel modified tobacco products. However the seat belts involve no endorsement of unsafe driving, while the modified tobacco products, if the government allows regulated claims of reduced harms, does provide endorsement of the very act of smoking tobacco. This heterogeneity complicates projections from the other harm reduction experiences to PREPs. However it indicates where one might turn for insight into likely effects. CONCLUSIONS Harm reduction is a viable approach to government interventions. It has informed policy debates in a number of areas, and there is a modest research base indicating that some interventions with a harm reduction focus may indeed be harm-reducing. However, the framework is not yet well developed in either theory or application and continues to encounter both popular and professional skepticism. That skepticism will form part of the backdrop to decisions about implementation of the framework in the area of reducing tobacco-related harms. Social Values There are well-documented health effects of tobacco exposure on the nonuser. Environmental tobacco smoke has just been included on the National Toxicology Program’s list of known-carcinogens and has long been linked to respiratory diseases and cancers in nontobacco users exposed to the smoke of others. Chapter 15 includes well-documented

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction evidence of effects of tobacco exposure in utero. One could argue that these involuntarily exposed people should be the touchstone for harm reduction demonstration. However it is difficult to identify other applications of the harm reduction framework in which this kind of reasoning has been used; for example, automobile safety modifications may have shifted the burden of traffic accidents to non-occupants, but this has not been given prominence. Another important consideration in decision making about harm reduction is the social value of the product. Many judgments about regulating risks include such considerations. For example, some risks of pharmaceutical products are acceptable because the benefit outweighs the risk, with benefit measured as decreased mortality, decreased disease prevalence or severity, and in some cases as quality-of-life and social well being. As a society, we accept certain levels of air pollution and attendant harm to health because we value the right to drive a car. Freedom of choice (to smoke or not) and the tobacco industry’s right to exist are values to the American public that preclude tobacco prohibition. However, the American public overwhelmingly supports restrictions or bans on tobacco use in indoor public places (The Gallup Organization, 2000) and restriction of youth access to tobacco products (American Heart Association, 1998). Moreover, most tobacco users themselves would like to quit and there are very few health benefits associated with tobacco use (See Chapter 16). Therefore, acceptable restrictions on potential exposure or harm reduction products might be stringent, although not severe enough to put the industry out of business. However, just as social values assure the continued availability of tobacco, social values can influence the limits of harm reduction strategies. It is unclear how much actual reduction in harm should be required for approval and marketing of a harm reduction product. Should a new product be allowed and encouraged if the gain is incremental only? Because regulation of novel tobacco or cigarette-like products involves a change in philosophy about tobacco, and because this is fundamentally a value judgment absent specific scientific guidance, one gauge of how much harm reduction should be required of a novel tobacco or cigarette-like product could be an assessment of how the public values increased health or reduced harm from tobacco. Such information is obtainable in a scientific way and could guide policymakers as they set the “bar” for regulatory action. Analytic Problems Though conceptually simple, harm reduction is difficult to apply when assessing alternative interventions. Following MacCoun and Reuter (2001), three basic problems emerge:

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction Aggregating harms. For illicit drugs in particular, a very heterogeneous set of adverse consequences need to be considered, such as loss of family cohesion, the spread of HIV and crime. Even if each effect can be measured, itself a major undertaking for any nonmarginal intervention (Reuter, 1999), there is little agreement about how to value them in monetary terms. The range of adverse consequences is narrower for the other cases, but even for alcohol, the range is wide enough (including crime and health) that aggregate measures in monetary terms are very approximate (e.g., Harwood et al, 1998). Some of these consequences are difficult to estimate and may consequently have been set at zero in quantitative studies because there is no systematic empirical base for an estimate. The harms derived from tobacco-related products are predominantly health related but the debate around economic interests of farmers in tobacco policy is a reminder that some other outcomes need to be taken into account. Weighting the interests of different parties. The interests of users versus non-users has already been mentioned but there may also be other considerations of relative culpability and vulnerability. For example, the health burden of the poor may be given greater weight or consideration than the burden borne by the nonpoor, simply because the former can do less to ameliorate their own problems. It is also important to note that harms are not uniformly distributed; any particular option may benefit one group at the expense of another. Harm reduction as a hazardous policy choice. An unsuccessful harm reduction intervention may lead to long-lasting and broadly distributed adverse consequences. For example, methadone maintenance might have turned out to increase heroin initiation and to prolong opiate addiction; it would be difficult to reverse those consequences or to be able to predict them in the early years of the intervention, let alone in advance of the decision. That suggests that harm reduction interventions may have to be held to a higher standard of proof and that government should be particularly careful, adopting one step at a time and closely monitoring the results. The fact that it will take decades to be certain about harm reduction effects of tobacco-related PREPs is a reason for particular caution in this case; this holds for none of the other applications. MacCoun (1998) summarizes the research on risk compensation: “[t]o date, research on compensatory responses to risk reduction provides little evidence that behavioral responses produce net increases in harm, or even the constant level of harm predicted by the “homeostatic” version of the theory. Instead, most studies find that when programs reduce the probability of harm given unsafe conduct, any increases in the probability of that conduct are slight, reducing, but not eliminating, the gains in safety” (Chirinko and Harper, 1993; Stetzer and Hofman, 1996). However, the

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction applicability of this finding to specific tobacco interventions remains to be established. The committee concludes with two other observations: Harm reduction presents major problems of measurement. Harm is much more difficult to measure than is prevalence or even quantity consumed, the conventional targets of control. As Single (1997) notes with respect to illegal drugs: “[A]s a practical matter, it is very difficult to determine whether specific policies involve a net reduction in drug-related harm.” It requires estimation not only of numerous disparate outcomes but also assessment of how much of their change can be attributed to the intervention. For example, vehicle fatalities are determined by many factors; it is a complex research task to identify the contribution of seat belt laws or more stringent Blood Alcohol Content laws. Harm reduction in the tobacco control field will require the development of complex surveillance programs and very difficult issues of attribution. Public health advocates opine that tobacco is a “special case,” because tobacco is the only consumer product that when used exactly as intended is lethal. Further, they posit that it is unconscionable to market an addictive product to youth who are not competent to make informed judgments about long-term risks in the face of perceived short-term benefits. Finally, an undeniable history of suppression of information about the health risks of tobacco and tobacco product design changes leads many to seriously question any assessment of harm reduction potential by the manufacturers of the products. Just as harm reduction with respect to illicit drugs has been hurt by its association with the legalization movement, so too has the tobacco companies’ use of false messages about the benefits of light and filter-tipped cigarettes created suspicion in the field of tobacco control. Therefore, the burden of proof for a benefit of novel, potential exposure or harm reduction tobacco products entails special considerations beyond that required of many other scientific questions. REFERENCES American Heart Association. 1998. Tobacco industry’s political and economic influence. [Online]. Available: http://www.americanheart.org/Heart_and_Stroke_A_Zguide/tobec.html [accessed 2001]. Anglin MD, Brecht ML, Woodward JA, et al. 1986. An empirical study of maturing out: conditional factors. International Journal of Addiction 21:233–246. Ball JC, Ross A. 1991. The Effectiveness of Methadone Maintenance: Patients, Programs, Services and Outcomes. New York: Praeger. Caulkins JP, Reuter P. 1997. Setting goals for drug policy: harm reduction or use reduction? Addiction 92(9):1143–1150.

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction CDC (Centers for Disease Control and Prevention). 1995. Update: alcohol-related traffic crashes and fatalities among youth and young adults-United States, 1982–1994. Morbidity and Mortality Weekly Report 44(47):869–874. Chirinko RS, Harper EP. 1993. Buckle up or slow down? New estimates of offsetting behavior and their implications for automobile safety regulation. Journal of Policy Analysis and Management 12(2):270–296. Crandall RW, Graham JD. 1989. The effect of fuel economy standards on automobile safety. Journal of Law and Economics 32(1):97–118. Dee TS. 1998. Reconsidering the effects of seat belt laws and their enforcement status. Accid Anal Prev 30(1):1–10. Duncan DF. 1997. Chronic drinking, binge drinking and drunk driving. Psychological Reports 80:681–682. Edwards G. 1995. Alcohol Policy and the Public Good. Oxford: Oxford University Press. Evans L. 1991. Traffic Safety and the Driver. New York: Van Nostrand Reinhold. Federal Interagency Forum on Child and Family Statistics. 1997. America’s Children: Key National Indicators of Well-Being. Washington, DC: U.S. Government Printing Office and the U.S. Department of Health and Human Services. Foy DW, Nunn LB, Rychtarik RG. 1984. Broad-spectrum behavioral treatment for chronic alcoholics: effects of trained controlled drinking skills. J Consult Clin Psychol 52(2):218– 230. Furstenberg FF Jr, Geitz LM, Teitler JO, Weiss CC. 1997. Does condom availability make a difference? An evaluation of Philadelphia’s health resource centers. Fam Plann Perspect 29(3):123–7. Guttmacher S, Lieberman L, Ward D, Freudenberg N, Radosh A, Des Jarlais D. 1997. Condom availability in New York City public high schools: relationships to condom use and sexual behavior. Am J Public Health 87(9):1427–33. Harwood H, Fountain D, Livermore G. 1998. The Economic Costs of Alcohol and Drug Abuse in the United States, 1992. Rockville, MD: National Institute on Drug Abuse. Henry J.Kaiser Family Foundation. 1996. The Kaiser Survey on Americans and AIDS/HIV. Menlo Park, CA: Kaiser Family Foundation. Henshaw SK. 1999. U.S. Teen Pregnancy Statistics: With Comparative Statistics for Women Aged 20–24. New York: Alan Guttmacher Institute. IOM (Institute of Medicine). 1994. Lynch BS, Bonnie RJ, eds. Growing Up Tobacco Free. Washington, DC: National Academy Press. IOM (Institute of Medicine). 1998. Taking Action to Reduce Tobacco Use. Washington, DC: National Academy Press. IOM (Institute of Medicine). 2000. No Time to Lose: Getting More From HIV Prevention. Washington, DC: National Academy Press. Kazman S. 1991. Death by regulation. Regulation 14(4):18–22. Kirby D. 1992. School-based programs to reduce sexual risk-taking behaviors. J Sch Health 62(7):280–7. Kirby D. 1997. No Easy Answers: Research Findings on Programs to Reduce Teen Pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy. Kirby DB, Brown NL. 1996. Condom availability programs in U.S. schools. Fam Plann Perspect 28(5):196–202. Kirby D, Brener ND, Brown NL, Peterfreund N, Hillard P, Harrist R. 1999. The impact of condom availability [correction of distribution] in Seattle schools on sexual behavior and condom use [published erratum appears in Am J Public Health. 1999; 89(3):422]. Am J Public Health 89(2):182–7. Kirby D, Coyle K. 1997. School-based programs to reduce sexual risk-taking behavior. Children and Youth Services Review 19(5–6):415–436.

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction Kreitman N. 1986. Alcohol consumption and the preventive paradox [see comments]. Br J Addict 81(3):353–63. Lenton S, Single E. 1998. The definition of harm reduction. Drug and Alcohol Review 17(2):213. Longshore D, Hsieh S, Danila B, Anglin MD. 1993. Methadone maintenance and needle/ syringe sharing. Int J Addict 28(10):983–96. MacCoun RJ. 1998. Toward a psychology of harm reduction. Am Psychol 53(11):1199–208. MacCoun RJ, Reuter P. 2001. Drug War Heresies: Learning From Other Vices, Times and Places. Cambridge, England: Cambridge University Press. MacCoun R, Reuter P, Schelling T. 1996. Assessing alternative drug control regimes. J Policy Anal Manage 15(3):330–52. Marlatt GA, Larimer ME, Baer JS, Quigley LA. 1993. Harm reduction for alcohol problems: Moving beyond the controlled drinking controversy. Behavior Therapy 24(4):461–504. Miller WR, Caddy GR. 1977. Abstinence and controlled drinking in the treatment of problem drinkers. Journal of Studies on Alcohol 38:986–1003. National Safety Council. 1999. Report on injuries in America. [Online]. Available: http://www.nsc.org/lrs/statinfo/99report.htm [accessed October 19, 2000]. Newman RG, Peyser N. 1991. Methadone treatment: experiment and experience. Journal of Psycoactive Drugs 23:115–121. NHTSA (National Highway Traffic Safety Administration). 1999a. Traffic safety facts 1998: speeding. [Online]. Available: http://www.nhtsa.dot.gov/people/ncsa/pdf/Speeding98.pdf [accessed October 19, 2000]. NHTSA (National Highway Traffic Safety Administration). 1999b. Traffic safety facts 1998: occupant protection. [Online]. Available: http://www.nhtsa.dot.gov/people/ncsa/pdf/OccPrt98.pdf [accessed October 19, 2000]. NIAAA (National Institute on Alcohol Abuse and Alcoholism). 2000. Tenth Special Report to the U.S. Congress on Alcohol and Health. Washington, DC: U.S. Department of Health and Human Services and National Institute on Alcohol Abuse and Alcoholism. Normand J, Vlahov D, Moses L. 1995. Preventing HIV Transmission: The Role of Sterile Needles and Bleach. Washington, DC: National Academy Press. Peltzmann S. 1975. The effects of automobile safety regulation. Journal of Political Economy 83(4):677–725. Plant M, Miller P, Nichol P. 1994. Preventing injuries from bar glasses-no such thing as safe glass. British Medical Journal 308(6938):1237–1238. Rettig R, Yarmolinsky A. 1995. Federal Regulation of Methadone Treatment. Washington, DC: National Academy Press. Reuter P. 1999. Are calculations of the economic costs of drug abuse either possible or useful? [comment]. Addiction 94(5):635–8. Rosenbaum M. 1997. The De-medicalization of methadone. Erickson PG, Riley DM, Cheung YW, O’Hare PA, ed. Harm Reduction: A New Direction for Drug Policies and Programs. Toronto, Canada: University of Toronto Press. Pp. 69–79. Ross HL. 1992. Controlling Drug Driving: Social Policy for Saving Lives. New Haven: Yale University Press. Single E. 1997. The concept of harm reduction and its application to alcohol: The 6th Dorothy Black lecture. Drugs: Education, Prevention and Policy 4(1):7–22. Single E, Rohl T. 1997. The National Drug Strategy: Mapping the Future. Canberra, AGPS: Report Commissioned by the Ministril Council on Drug Stategy. Sobell MB, Sobell LC. 1978. Behavioral Treatment of Alcohol Problems: Individualized Therapy and Controlled Drinking. New York: Plenum Press. Stetzer A, Hofman DA. 1996. Risk compensation: implications for safety interventions. Organizational Behavior and Human Decision Processes 66(1):73–88.

OCR for page 38
Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction Stockwell T, Hawks D, Lang E, Rydon P. 1996. Unravelling the preventive paradox for acute alcohol problems. Drug and Alcohol Review 15(1):7–15. Tengs TO, Adams ME, Pliskin JS, et al. 1995. Five-hundred life-saving interventions and their cost-effectiveness [see comments]. Risk Anal 15(3):369–390. The Gallup Organization. November 29, 2000. Smoking in restaurants frowned on by many Americans. [Online]. Available: http://www.gallup.com/poll/releases/pr001129.asp [accessed 2001]. U.S. DHHS (U.S. Department of Health and Human Services). April 20, 1998. Press Release: Research shows needle exchange programs reduce HIV infections without increasing drug use. [Online]. Available: http://www.hhs.gov/news/press/1998pres/980420a.html [accessed October 19, 2000]. U.S. DHHS (U.S. Department of Health and Human Services). 1994. Preventing Tobacco Use Among Young People; A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. U.S. DHHS (U.S. Department of Health and Human Services). 1988. The Health Consequences of Smoking; Nicotine Addiction; A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Vaillant G. 1996. A long-term follow-up of male alcohol abuse. Archives of General Psychiatry 53:243–249. Ventura SJ, Curtin SC, Mathews TJ. 1997. Teenage Births in the United States: National and State Trends, 1990–96. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Health Statistics. Ward J, Mattick R, Hall W. 1992. Key Issues in Methadone Maintenance Treatment. Australia: University of New South Wales Press. White House Conerence for a Drug Free America (1988). Wilde GJS. 1982. The theory of risk homeostasis: implications for safety and health. Risk Analysis 2:209–255.