No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Using drug courts for drug postmarketing surveillance
Abstract
Purpose
– One of the greatest challenges for drug regulation is valid, comprehensive surveillance of drugs after they reach the pharmaceutical market. The purpose of this paper is to propose a new method of individual and aggregate-level postmarket surveillance using data previously (and continuously) collected by drug courts, which are in operation in nearly every geographic corner of the USA.
Design/methodology/approach
– To determine the feasibility of such an undertaking, data were obtained from an urban, southern county drug court. Intake data included all participants from September 2012 to November 2013. The final sample included 532 drug court participants.
Findings
– Intake data were found to include various demographic variables, measures of drug use, and various sociological/criminological variables such as familial and social support, church attendance, and other pertinent variables for studying drug use and crime trends generally.
Practical implications
– By using intake data from drug courts in a manner similar to Uniform Crime Report or National Incident-Based Reporting System, this could add greatly to the understanding of crime and drug use.
Social implications
– The authors purport that a data management system of drug court intake data could provide a cost-efficient and generalizable representation of drug use of those within the criminal justice system.
Originality/value
– Many efforts have been employed in an attempt to better ascertain where high rates of drug use occur. By using drug courts as more than just a system of treatment, postmarketing surveillance could be improved.
... However, the expansion and growth in the number of drug courts has been remarkable, especially when compared to many other criminal justice programs. In 2014, by some estimates, there were nearly 3,000 drug courts in existence in the United States (Griffin & Woodward, 2016). While there is some evidence that the growth of drug courts is attributable to the effectiveness of these programs, one economic reason that so many drug courts have been established is that the federal government has allocated considerable financial resources for jurisdictions that are interested in establishing these programs (Jensen & Mosher, 2006;Listwan, Sundt, Holsinger, & Latessa, 2003). ...
One method that has been touted to help end mass incarceration is using intermediate sanctions. While intermediate sanctions often present as attractive options, there is evidence that as practiced, these sanctions often result in net widening. One of the most common forms of intermediate sanctions are drug courts, which are often viewed as progressive alternatives to locking up people with substance abuse problems. However, along with the dangers of net widening, scholars have shown that many people admitted to drug courts do not seem to have substance abuse problems and could benefit from lesser criminal justice interventions. In the current study, we analyzed intake data from a drug court to determine: (1) what charge(s) drug participants had and (2) how they became involved with the criminal justice system. Among important findings were that a large number of drug court participants were arrested for the possession of one drug only (often marijuana) and that more than half of participants came to the attention of the criminal justice system through a traffic stop rather than through repeated encounters with the criminal justice system.
The goal of this chapter is to provide an overview of the recently reju-venated field of comparative criminology. It begins by considering the context and history of comparative criminology and continues by out-lining the contemporary comparative perspective. After identifying several goals for comparative criminology that are often advanced, including theory elaboration and testing as well as policy evaluation and critique, the chapter describes the common approaches to com-parative criminological research. The main theoretical traditions of comparative criminology are examined first, with particular attention directed to metanarratives such as modernization, civilization, oppor-tunity, and world system theories and to structural theories based on culture, social bonds, and the distribution of economic resources. Taking up methodological concerns next, the chapter summarizes some of the more common dependent variables studied by compara-tive criminologists, noting how these variables have been operational-ized in the literature, then explores the three methodological approaches most typically deployed in the field, specifically metalev-el, parallel, and case studies.
In this paper, we examine the relationship between drug use and gang member-ship using data from the Arizona Arrestee Drug Abuse Monitoring (ADAM) program, which collects both self-report and hard measures (i.e., urinalysis) of drug use. Our analyses revealed that self-reported recent drug use (i.e., drug use in the past three days) and urinalysis outcomes were similarly associated with the gang-membership variables. These findings suggest that self-reported data obtained from gang members is a particularly robust method for gathering information on their recent behavior. Additionally, our results were supportive of the social facilitation model, showing that current gang members were signif-icantly more likely to use marijuana and cocaine compared with former gang members. The implications for policy and future research are discussed.
Data collected in the Arrestee Drug Abuse Monitoring (ADAM) program on alcohol and other drug use among arrestees provide a valuable opportunity to examine the relationship between alcohol use and violence. The data are used to explore the combined use of alcohol and other drugs among offenders and the relationships between substance use and the offenders' demographic characteristics and offenses. These findings are used to identify changes in the offenders' alcohol and other drug use over time.
OxyContin, a controlled-release opioid developed and produced by Purdue Pharma, was given Food and Drug Administration (FDA) approval in December 1995. By 1999, to the apparent surprise of Purdue Pharma, the Drug Enforcement Administration (DEA), and the FDA, extensive reports of OxyContin abuse and diversion began to circulate. The drug abuse assessment liability system in the United States has often been criticized and the experience with OxyContin did nothing but buttress those criticisms. However, as seven investigational hearings conducted by the U. S. Congress would expose, several changes were made to the abuse liability assessment to correct deficiencies in the system which had aggravated OxyContin abuse and diversion-a welcome sign from what many have considered an American tragedy.
This research, drawing on social disorganization and social deregulation theories, examines the effects of static and dynamic measures of the ecological structure of municipalities on the change in the level of robbery and homicide rates during the 1970s. The data analyses, for a sample of 109 U.S. cities, reveal that the change in the level of crime is affected by the change in the ecological structure of cities, as well as ongoing social processes. We interpret these findings as indicating that structural conditions which weaken social cohesion produce an increase in the level of crime.
The NIBRS data program currently being implemented by the FBI and local law enforcement agencies has by now produced sufficient data for archiving and distribution. Although not representative of crime in the United States, existing NIBRS data can be used to investigate the nature of crimes known to the police compared to the traditional UCR data. The Bureau of Justice Statistics has requested the National Archive of Criminal Justice Data to store and make NIBRS data available to interested users. The data from 1996 will shortly be available from the NACJD web site. The 1996 data contain almost 6.5 million records and the FBI’s full file includes about 361 Mbytes of data. The data have been disaggregated from the FBI’s complex single file into 11 segment levels or record types. This makes the individual record types easier and faster to analyze than using the full file, which more closely resembles a relational database than a hierarchical file. However, splitting apart the record types requires that special procedures be used to merge files of different record types, which would be necessary if a user were interested in analyzing variables appearing in more than one record type (e.g., comparing offender and victim ages). These procedures are described, and a test comparing the time to run a simple frequency count using the full file against the merged files shows that using the merged files is considerably more efficient. Also discussed are some future developments to facilitate the analysis of NIBRS data.
The globalization of biopharmaceutical clinical trials and their offshore outsourcing, from the West to low and middle-income countries, has come under increasing scrutiny from academic scholars, practitioners, regulatory agencies and the media. This article reports the results of a study conducted in Bangalore and Hyderabad between 2007-2009, to elicit the perspectives of stakeholders, concerning media representations of their work and the ethical issues that emanate from their engagement in the clinical trials enterprise. In acknowledging the inherently problematic nature of the outsourcing of clinical trials to low income countries, I argue that the practice of not prioritizing research on diseases that are most prevalent among communities, from which subjects are recruited, demands a coordinated and sustained critique. I propose that the critical discourse on the outsourcing of clinical trials should not only emphasize the perils of this practice, but also address some broader issues of equity and distributive justice that determine people’s access to basic health care in low income countries. Close attention to the specific context of clinical trials in an increasingly neoliberal medical and health environment in emerging economies such as India can provide critical insights into the on-the-ground complexities and challenges of outsourced global clinical trials.
We use data from the National Incident-Based Reporting System to examine the effects of offender and victim age on whether male offenders commit sexual assault while robbing women. Restricting analyses to robberies reveals the offenders' age preferences since it allows one to control for the effects of opportunity. We find that robbers of all ages are most likely to sexually assault women at ages 15–29 years, ages when their reproductive potential is highest. However, in contrast to the idea that rape is a direct adaptation, victims are no more likely to be raped than sexually assaulted at these ages. The age of the offender is also a strong predictor of sexual assault. The likelihood that a robber commits a sexual assault increases from age 12 years until he reaches his early thirties when it begins to decline. This age pattern corresponds, to some extent, to age differences in the male sex drive.
Although research has shown specific land uses to be related to crime, systematic investigation of land uses and violent crime has been less common. This study systematically examines links between land uses and violent crime and assesses whether such links are conditioned by socioeconomic disadvantage. We employ geocoded Uniform Crime Report (UCR) data from the Indianapolis police department and information on 30 categories of land use and demographic information from the 2000 U.S. Census. We use land use variables to predict violent crime counts in 1,000 × 1,000-feet grid cells using negative binomial regression models. Results indicate that, net of other variables, specific land uses predict variation in counts for individual violent crimes and aggregate rates. Some nonresidential land uses are associated with higher violent crime counts, whereas others are associated with lower counts. Specific land uses also condition the effects of socioeconomic disadvantage on violent crime. The implications for routine activity/opportunity and social disorganization/collective efficacy theories of crime are discussed.
The National Incident-Based Reporting System (NIBRS) offers a new source ofdata for measuring crime. Though far from national in coverage, NIBRS datacan address research and policy questions outside the scope of the UniformCrime Reports and the National Crime Victimization Survey. Comparisons ofthe three sources of crime data are presented, with particular emphasis onwhat can be learned from incident-based police data that cannot be learnedfrom other sources. Like all data on crime, the NIBRS is subject to variousproblems with validity and measurement error. Two general categories of suchproblems are discussed: those evident in the design of the NIBRS and thoselinked to more general issues in the organizational production of data.
Tramadol HCl, marketed as Ultram in the USA, was introduced as a non-scheduled drug in April 1995 based on the assumption that the risk of abuse was sufficiently low to warrant a non-scheduled status. However, approval was contingent upon the development of an innovative proactive surveillance program, to be overseen by an independent steering committee, which would detect unexpectedly high levels of abuse. The postmarketing surveillance program consisted of systematic collection and scientific evaluation of reports of suspected abuse in high-risk populations surveyed through an extensive key informant network of drug abuse specialists and all spontaneous reports of abuse received through the FDA MedWatch system. Methods to estimate the number of patients prescribed tramadol were also developed. Monthly rates of abuse were calculated as an index of the risk-benefit ratio (i.e., abuse cases per 100,000 patients prescribed the drug). The data for the 3 years since the drug was introduced show that the reported rate of abuse has been low. Although a period of experimentation seemed to occur in the first 18 months after its introduction--which reached a peak rate of approximately two cases per 100,000 patients exposed--during the 2 year period prior to June 1998, the reported rate of abuse has significantly (P = 0.011) declined, reaching levels of less than one case per 100,000 patients in the last 18 months. The overwhelming majority of abuse cases (97%) have been found to occur among individuals with a history of substance abuse and the abuse has been confined to isolated pockets around the country-notably none of which have significant populations of street drug abusers. Thus, the data support the decision not to schedule tramadol and, furthermore, suggest that a proactive post-marketing surveillance program can be successfully developed to effectively monitor abuse of new medications.
Tramadol has been marketed in the US since 1995. The US Food and Drug Administration agreed to release tramadol as a non-scheduled drug if proactive post-marketing surveillance studies would be conducted. This study was one of two phase IV protocols that were part of the overall surveillance program. It focused on impaired health professionals who are a high risk/high access population for drug abuse. All active participants in four state monitoring programs between November 1, 1995 and August 15, 1998 (n = 1,601) were recruited for the study. With the exceptions of implementing a standardized intake interview and urine testing for tramadol metabolites, all states operated their programs in the usual fashion. The programs were alerted to persistent non-prescribed tramadol use so that appropriate interventions could be employed. Despite availability of tramadol and the conditions that might lead to its abuse, the incidence rate for tramadol use in the study population was only 69 per thousand persons per year and the incidence rate for tramadol abuse or dependence was 6.9 per thousand persons per year.
The abuse liability of medications is a growing concern as the number of newly approved psychoactive medications increases. Postmarketing surveillance can assist in determining abuse liability, but strategies are not well-defined for medications believed to be at low abuse risk. Using a newly approved medication (sibutramine--an anorectic drug), a novel approach to postmarketing abuse surveillance was introduced. A one-page anonymous questionnaire covering sibutramine, a scheduled anorectic drug (phentermine), and a fabricated name was added to the intake process of 58 treatment programs. From the 8780 completed questionnaires, 8.8% had heard of sibutramine and phentermine. For continued use to get high (a proxy for abuse), the rate for sibutramine was lower than for phentermine (0.6 vs. 2.2%, McNemar's chi(2) = 110.45, P < 0.001) but was higher than for the fabricated name (0.6 vs. 0.3%, McNemar's chi(2) = 11.86, P < 0.001). These results suggest the risk of abuse associated with sibutramine was lower than that associated with a known abused drug, one that itself is considered low risk despite decades of population exposure. The relatively high rate of hearing of sibutramine may be due to the direct-to-consumer advertisement. This approach is only one indicator in a surveillance framework but appears promising and validates findings from laboratory-based abuse liability studies that also indicate low abuse liability for sibutramine.
Assessing actual abuse of prescribed medications requires postmarketing surveillance. In this article we discuss general systems of postmarketing surveillance that exist as of the end of 2002 in the United States and two medication-specific surveillance systems that were devised and tested. The two specific surveillance systems are compared with limitations highlighted. Postmarketing surveillance is in its infancy and requires more research on ways to improve its validity without inducing illicit experimentation. Information on comparator medications is highly recommended both to validate the system and to place the results in context.
In this article, the use of data collected in registration-focused clinical trials to provide information concerning abuse liability of compounds under development is discussed. Registration-focused trials are limited by the small and select sample chosen for study participation and design constraints. However, most compounds under development are first administered to humans during the conduct of registration-focused trials, so this presents an opportunity to collect potentially important information about the effects of drugs in humans. At present, information concerning subjective effects and symptoms associated with drug discontinuation are not collected systematically. Reports are generally considered as adverse events (AEs) and are recorded on the case report forms (CRFs) in the investigators own language. There is generally no rating of drug "liking". The authors suggest strategies that could be implemented in registration-focused clinical trials to improve the information gathered with regard to subjective effects, abuse liability and discontinuation-emergent symptoms. Importantly, there remains much groundwork to be done in developing and validating appropriate assessment instruments and determining "threshold" levels for concern. Under the best of circumstances, registration-focused clinical trials have limited potential to detect abuse liability because of the small number of patients seen and the exclusion of many subjects who might be particularly vulnerable to the abuse of marketed compounds (i.e. individuals with substance use disorders). In cases where there are reasons to suspect that a drug under development has abuse potential, systematic exploration with a series of studies specifically designed to assess abuse potential must be conducted.
This paper describes the rationale for use of preclinical assessments of abuse liability in laboratory animals, and then discusses "cross-cutting" methodological issues that apply to behavioral evaluations intended to contribute to an abuse liability evaluation package. Issues include use of: (1) positive and negative control conditions; (2) full dose-effect evaluations, (3) multiple dependent measures, (4) pharmacokinetic evaluations to guide choice of dose ranges, (5) a species for which good methodological and comparative data are available to aid interpretation of results, and (6) appropriate methods for the group or single-subject experimental design selected. The remainder of the paper describes basic methodology by which three core pieces of behavioral data required by the Food and Drug Administration for its use in the overall abuse liability analysis can be obtained preclinically. Reinforcing effects are assessed in study of drug self-administration; drug discrimination assesses degree of overlap of interoceptive stimulus effects with relevant comparison drugs; physical dependence potential is determined by assessing whether a withdrawal syndrome occurs after chronic drug administration. Background and methodological issues specific to each procedure are discussed. A key consideration for cross-cutting and specific methodological issues is that choices made enable confident interpretation of both positive and negative results.
This paper describes the rationale and procedures for conducting what is considered by many to be the current "gold standard" for initial abuse liability testing of a novel compound: the classic acute dose-effect comparison study in volunteers with histories of drug abuse. Such a trial is most appropriate for predicting the likelihood of abuse by drug abusers and, in turn, the extent of drug diversion and illicit street sales if the novel compound became available in the community. The dose-effect abuse liability trial typically involves a double-blind complete crossover design in 10-14 subjects with histories of polydrug abuse in a controlled clinical pharmacology laboratory setting. Drug conditions usually involve placebo, three doses of the novel compound and three doses of an appropriate reference compound of known abuse liability. In each session, the time-course of effects of a single drug dose are evaluated. Intervals between experimental sessions are typically 1 to several days. The importance of testing high supra-therapeutic doses of the novel drug for the validity of the trial is emphasized, and the use of a dose run-up pilot study for selecting maximal doses and matching doses between the novel and comparison compound is explained. The rationale and description of outcome measures is discussed, including measures that reflect likelihood of abuse (e.g. drug vs. money choice and subject ratings of liking, good effects, estimated monetary street value), secondary measures that should be considered in interpreting likelihood of abuse (e.g. drug identification, subject-rated side effects and mood changes), and additional concurrent measures to establish equivalence of the novel and comparison compound (e.g. behavioral performance, observer-rated assessments, physiological measures).
Regulatory control of drugs with abuse liability is an important component of drug control policy and is believed to help prevent nonmedical use. To be maximally effective, this requires a scientific assessment of abuse liability of drugs considered for regulatory control. These assessments have relied extensively on laboratory-based animal and human testing, but also utilize information from clinical trials, actual abuse and other sources. Here, we discuss recommendations and guidelines that have been proposed for abuse liability assessment and describe important review papers and conference proceedings that have addressed this matter, focusing primarily on drugs with medical usefulness. Historically, there is substantial consensus about how to approach abuse liability evaluation of drugs with actions similar to those of abused opiates, stimulants, depressants, and to a somewhat lesser extent, cannabinoids and hallucinogens, and much of what has been recommended for abuse potential assessment in the past remains valid and useful. On the other hand, novel CNS-active medications which cannot be readily classified with these traditional drugs of abuse are increasingly under development. In addition, advances in the science of abuse liability assessment need to be incorporated into future guidelines and recommendations on this subject. Developers of new medications need guidance on how to utilize scientific research to maximize therapeutic benefit while minimizing risk for abuse. Thus, another goal of this review has been to identify areas where critical thinking and new guideline development are needed.
To demonstrate the utility of postmarketing studies using in-treatment drug and alcohol abusers as informants for assessing the relative abuse liability of sedative-hypnotic drugs.
A survey was conducted that ascertained exposure to a variety of drugs with hypnotic/sedative properties and elicited subjective evaluations indicative of abuse liability.
Complete data were obtained from 297 admissions (78% male) to three addiction treatment sites in the United Kingdom. Subjects were asked 15 questions about 12 different drugs, including five benzodiazepines, three antidepressants, two non-benzodiazepine hypnotics and two antihistamines (plus one fictitious drug). Three of the benzodiazepines (diazepam, nitrazipam, temazepam) emerged as having substantially more abuse liability than any of the other drugs tested, as revealed by higher scores on abuse liability items (purchased on street, taken to get high, like drug, potential for addiction to drug). The antihistamines (chlorpheniramine, diphenhydramine) had lowest abuse liability profiles, while the antidepressants (amitriptyline, fluoxetine, trazadone) and non-benzodiazepine hypnotics (zolpidem, zopiclone) had similar profiles.
This pilot study suggests that postmarketing information on hypnotic drug use obtained from drug addicts entering treatment produces data consistent with other measures of abuse liability. The data suggest that the risk of misuse of newer non-benzodiazepine hypnotics may be less than that of benzodiazepine drugs, and similar to that of sedating antidepressants. The new methodology may serve to clarify or validate premarketing abuse liability data, and may help to inform the regulatory process and physician practice.
Scheduling of a chemical drug substance under the Controlled Substances Act (CSA) includes an evaluation of preclinical and clinical safety, and experimental abuse liability studies, as well as information on diversion and overdose. Formulations that mitigate abuse liability, dependence potential and public health risks (e.g., altered absorption rate and tamperability, long half-life, pro-drugs and combination products) are amenable to preclinical and clinical studies to compare their abuse potential to reference compounds. For new formulations (NF) as marketed agents, direct comparison to the immediate release (IR) formulation of the reference compound is typically needed across the full range of potential studies. While the public health advantage of formulation changes in the marketplace can be conceptualized in behavioral economic terms, generating persuasive data is challenging. Study complexity increases because of additional conditions (e.g., placebo, 2-3 doses of the IR formulation, 2-3 doses of the new formulation, and 2-3 doses of the unscheduled or negative control drug), larger sample sizes (study power driven by the comparison of the new formulation versus the IR or placebo), and associated increases in study duration. However, the use of single maximal doses of well-characterized controls can reduce the number of study arms, and using incomplete block designs can reduce study duration. Less typical experimental approaches may also be useful, such as human choice or discrimination procedures, or pre-marketing consumer studies among experienced drug tamperers. New formulations that demonstrate a substantial difference from marketed or reference products have a potential marketing advantage and should require less onerous risk management. Post-marketing epidemiological data demonstrating the lack of abuse will carry the most weight from a public health and physician perspective.
There is a demand for pharmaceutical products with reduced abuse liability. These products must meet three tests to be successful. They must be safe for patients, be less likely to injure the abuser, and be less desirable for abuse by established drug abusers relative to existing products on a dose for dose (milligram-equivalent) basis. There is a need for standardization of the evaluation of abusable pharmaceuticals in the various stages of drug development from preclinical animal studies to postmarketing surveillance. Formulations with reduced abuse liability must: (1) be tested using standard animal, benchtop, and human pharmacokinetic methods that allow interpretation, (2) sufficiently reduce the recovery of abusable drug substance, or contain another ingredient to deter abuse, (3) not alter drug activity for patients in an undesirable or risky way, and (4) have an accurate pre-approval estimation of their reduced abuse liability, which is validated by adequate epidemiologic post-approval surveillance.
Use and production of methamphetamine (MA) has dramatically increased in the United States, especially in rural areas, with concomitant burdens on the treatment and criminal justice systems. However, cocaine is also widely used in many rural areas. The purpose of this article is to contrast MA and cocaine users in three geographically distinct rural areas of the United States.
Participants were recent not-in-treatment adult cocaine and MA users living in rural Ohio, Arkansas, and Kentucky, who were recruited by a referral recruitment method for sampling hidden community populations. Participants were interviewed for demographics, drug and alcohol use, criminal justice involvement, and psychological distress (Brief Symptom Inventory).
The sample of 706 comprised 29% nonwhite and 38% female participants; the average age was 32.6 years; 58% had a high school education or higher, and 32% were employed. In the past 6 months, they had used either MA only (13%), cocaine only (52%), or both (35%). MA users were seldom (8.2%) nonwhite, but type of stimulant use did not vary by gender. Combined MA/cocaine users reported significantly greater use of alcohol and other drugs, including marijuana and nonprescribed opiates and tranquilizers, and reported significantly higher psychological distress. MA users (with or without cocaine use) had greater odds of recent criminal justice involvement compared with cocaine-only users.
There is a clear need for accessible substance-use treatment and prevention services in rural areas of the United States, including services that can address MA, cocaine, polydrug use, and mental health needs. There is a particular need of these services for polydrug users.
OBJECTIVE. Beginning in the late 1990's a marked increase in abuse of OxyContin emerged, which led to the development and establishment of a proactive surveillance program to monitor and characterize abuse, named the Researched Abuse, Diversion and Addiction Related Surveillance (RADARS) System. The main goal of RADARS was to develop proactive, timely and geographically sensitive methods to assess the abuse and diversion of OxyContin, along with a number of other Schedule II and III opioids with the aim of using this information to guide risk reduction interventions. Thus, its major focus was the detection of abuse of OxyContin and other commonly prescribed opioid analgesics at the three-digit ZIP code level across the country utilizing a number of different detection systems.
The detection systems selected were: (1) Quarterly-surveys of drug abuse experts who are knowledgeable about cases of prescription drug abuse; (2) Surveys of law enforcement agencies that detect diversion of prescription drugs; and (3) Poison Control Center reports of intentional misuse or abuse of prescription opioids. Collectively, the three systems provide overlapping coverage of over 80% of the nation's 973 three-digit ZIP codes.
Preliminary results indicate that prescription drug abuse is prevalent nationwide, but it seems to be heavily localized in rural, suburban and small urban areas. Our results also indicate that hydrocodone and extended and immediate release oxycodone products are by far the most widely abused drugs in the country, but the abuse of all prescription opioids seems to have grown over the 14 quarters since the inception of RADARS.
The next step in these studies is to develop regionally specific, risk-minimization-strategies, which is the goal of all risk-management programs. If successful, RADARS will serve as a prototype of such programs for any new drug approved that has measurable abuse potential.
Principles of initial experimental drug abuse liability assessment in humans
- R R Griffiths
- G E Bigelow
- N A Ator
The impact on criminal behavior and participant attitudes: results from NIJ’s multi-site adult drug court evaluation, part 2”, paper presented at National Association of Drug Court Professionals 16th Annual Training Conference
- M Rempel
- M Green
- K Henry
Inside the FDA: The Business and Politics Behind the Drugs We Take and the Food We Eat
- F Hawthorne
Hard drugs demanding solid understanding: column” USA Today
- B Kilmer
- J Caulkins
How many daily heroin users are there in the US? Somewhere between 60,000 and 1 million. Maybe”, Forbes: Opinion
- J Sullum